;; GCC machine description for IA-32 and x86-64.
;; Copyright (C) 1988, 1994, 1995, 1996, 1997, 1998, 1999, 2000,
;; 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009
;; Free Software Foundation, Inc.
;; Mostly by William Schelter.
;; x86_64 support added by Jan Hubicka
;;
;; This file is part of GCC.
;;
;; GCC is free software; you can redistribute it and/or modify
;; it under the terms of the GNU General Public License as published by
;; the Free Software Foundation; either version 3, or (at your option)
;; any later version.
;;
;; GCC is distributed in the hope that it will be useful,
;; but WITHOUT ANY WARRANTY; without even the implied warranty of
;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
;; GNU General Public License for more details.
;;
;; You should have received a copy of the GNU General Public License
;; along with GCC; see the file COPYING3. If not see
;; . */
;;
;; The original PO technology requires these to be ordered by speed,
;; so that assigner will pick the fastest.
;;
;; See file "rtl.def" for documentation on define_insn, match_*, et. al.
;;
;; The special asm out single letter directives following a '%' are:
;; L,W,B,Q,S,T -- print the opcode suffix for specified size of operand.
;; C -- print opcode suffix for set/cmov insn.
;; c -- like C, but print reversed condition
;; E,e -- likewise, but for compare-and-branch fused insn.
;; F,f -- likewise, but for floating-point.
;; O -- if HAVE_AS_IX86_CMOV_SUN_SYNTAX, expand to "w.", "l." or "q.",
;; otherwise nothing
;; R -- print the prefix for register names.
;; z -- print the opcode suffix for the size of the current operand.
;; Z -- likewise, with special suffixes for x87 instructions.
;; * -- print a star (in certain assembler syntax)
;; A -- print an absolute memory reference.
;; w -- print the operand as if it's a "word" (HImode) even if it isn't.
;; s -- print a shift double count, followed by the assemblers argument
;; delimiter.
;; b -- print the QImode name of the register for the indicated operand.
;; %b0 would print %al if operands[0] is reg 0.
;; w -- likewise, print the HImode name of the register.
;; k -- likewise, print the SImode name of the register.
;; q -- likewise, print the DImode name of the register.
;; x -- likewise, print the V4SFmode name of the register.
;; t -- likewise, print the V8SFmode name of the register.
;; h -- print the QImode name for a "high" register, either ah, bh, ch or dh.
;; y -- print "st(0)" instead of "st" as a register.
;; d -- print duplicated register operand for AVX instruction.
;; D -- print condition for SSE cmp instruction.
;; P -- if PIC, print an @PLT suffix.
;; X -- don't print any sort of PIC '@' suffix for a symbol.
;; & -- print some in-use local-dynamic symbol name.
;; H -- print a memory address offset by 8; used for sse high-parts
;; + -- print a branch hint as 'cs' or 'ds' prefix
;; ; -- print a semicolon (after prefixes due to bug in older gas).
;; UNSPEC usage:
(define_constants
[; Relocation specifiers
(UNSPEC_GOT 0)
(UNSPEC_GOTOFF 1)
(UNSPEC_GOTPCREL 2)
(UNSPEC_GOTTPOFF 3)
(UNSPEC_TPOFF 4)
(UNSPEC_NTPOFF 5)
(UNSPEC_DTPOFF 6)
(UNSPEC_GOTNTPOFF 7)
(UNSPEC_INDNTPOFF 8)
(UNSPEC_PLTOFF 9)
(UNSPEC_MACHOPIC_OFFSET 10)
; Prologue support
(UNSPEC_STACK_ALLOC 11)
(UNSPEC_SET_GOT 12)
(UNSPEC_SSE_PROLOGUE_SAVE 13)
(UNSPEC_REG_SAVE 14)
(UNSPEC_DEF_CFA 15)
(UNSPEC_SET_RIP 16)
(UNSPEC_SET_GOT_OFFSET 17)
(UNSPEC_MEMORY_BLOCKAGE 18)
; TLS support
(UNSPEC_TP 20)
(UNSPEC_TLS_GD 21)
(UNSPEC_TLS_LD_BASE 22)
(UNSPEC_TLSDESC 23)
; Other random patterns
(UNSPEC_SCAS 30)
(UNSPEC_FNSTSW 31)
(UNSPEC_SAHF 32)
(UNSPEC_FSTCW 33)
(UNSPEC_ADD_CARRY 34)
(UNSPEC_FLDCW 35)
(UNSPEC_REP 36)
(UNSPEC_LD_MPIC 38) ; load_macho_picbase
(UNSPEC_TRUNC_NOOP 39)
; For SSE/MMX support:
(UNSPEC_FIX_NOTRUNC 40)
(UNSPEC_MASKMOV 41)
(UNSPEC_MOVMSK 42)
(UNSPEC_MOVNT 43)
(UNSPEC_MOVU 44)
(UNSPEC_RCP 45)
(UNSPEC_RSQRT 46)
(UNSPEC_SFENCE 47)
(UNSPEC_PFRCP 49)
(UNSPEC_PFRCPIT1 40)
(UNSPEC_PFRCPIT2 41)
(UNSPEC_PFRSQRT 42)
(UNSPEC_PFRSQIT1 43)
(UNSPEC_MFENCE 44)
(UNSPEC_LFENCE 45)
(UNSPEC_PSADBW 46)
(UNSPEC_LDDQU 47)
(UNSPEC_MS_TO_SYSV_CALL 48)
; Generic math support
(UNSPEC_COPYSIGN 50)
(UNSPEC_IEEE_MIN 51) ; not commutative
(UNSPEC_IEEE_MAX 52) ; not commutative
; x87 Floating point
(UNSPEC_SIN 60)
(UNSPEC_COS 61)
(UNSPEC_FPATAN 62)
(UNSPEC_FYL2X 63)
(UNSPEC_FYL2XP1 64)
(UNSPEC_FRNDINT 65)
(UNSPEC_FIST 66)
(UNSPEC_F2XM1 67)
(UNSPEC_TAN 68)
(UNSPEC_FXAM 69)
; x87 Rounding
(UNSPEC_FRNDINT_FLOOR 70)
(UNSPEC_FRNDINT_CEIL 71)
(UNSPEC_FRNDINT_TRUNC 72)
(UNSPEC_FRNDINT_MASK_PM 73)
(UNSPEC_FIST_FLOOR 74)
(UNSPEC_FIST_CEIL 75)
; x87 Double output FP
(UNSPEC_SINCOS_COS 80)
(UNSPEC_SINCOS_SIN 81)
(UNSPEC_XTRACT_FRACT 84)
(UNSPEC_XTRACT_EXP 85)
(UNSPEC_FSCALE_FRACT 86)
(UNSPEC_FSCALE_EXP 87)
(UNSPEC_FPREM_F 88)
(UNSPEC_FPREM_U 89)
(UNSPEC_FPREM1_F 90)
(UNSPEC_FPREM1_U 91)
(UNSPEC_C2_FLAG 95)
(UNSPEC_FXAM_MEM 96)
; SSP patterns
(UNSPEC_SP_SET 100)
(UNSPEC_SP_TEST 101)
(UNSPEC_SP_TLS_SET 102)
(UNSPEC_SP_TLS_TEST 103)
; SSSE3
(UNSPEC_PSHUFB 120)
(UNSPEC_PSIGN 121)
(UNSPEC_PALIGNR 122)
; For SSE4A support
(UNSPEC_EXTRQI 130)
(UNSPEC_EXTRQ 131)
(UNSPEC_INSERTQI 132)
(UNSPEC_INSERTQ 133)
; For SSE4.1 support
(UNSPEC_BLENDV 134)
(UNSPEC_INSERTPS 135)
(UNSPEC_DP 136)
(UNSPEC_MOVNTDQA 137)
(UNSPEC_MPSADBW 138)
(UNSPEC_PHMINPOSUW 139)
(UNSPEC_PTEST 140)
(UNSPEC_ROUND 141)
; For SSE4.2 support
(UNSPEC_CRC32 143)
(UNSPEC_PCMPESTR 144)
(UNSPEC_PCMPISTR 145)
; For FMA4 support
(UNSPEC_FMA4_INTRINSIC 150)
(UNSPEC_FMA4_FMADDSUB 151)
(UNSPEC_FMA4_FMSUBADD 152)
; For AES support
(UNSPEC_AESENC 159)
(UNSPEC_AESENCLAST 160)
(UNSPEC_AESDEC 161)
(UNSPEC_AESDECLAST 162)
(UNSPEC_AESIMC 163)
(UNSPEC_AESKEYGENASSIST 164)
; For PCLMUL support
(UNSPEC_PCLMUL 165)
; For AVX support
(UNSPEC_PCMP 166)
(UNSPEC_VPERMIL 167)
(UNSPEC_VPERMIL2F128 168)
(UNSPEC_MASKLOAD 169)
(UNSPEC_MASKSTORE 170)
(UNSPEC_CAST 171)
(UNSPEC_VTESTP 172)
])
(define_constants
[(UNSPECV_BLOCKAGE 0)
(UNSPECV_STACK_PROBE 1)
(UNSPECV_EMMS 2)
(UNSPECV_LDMXCSR 3)
(UNSPECV_STMXCSR 4)
(UNSPECV_FEMMS 5)
(UNSPECV_CLFLUSH 6)
(UNSPECV_ALIGN 7)
(UNSPECV_MONITOR 8)
(UNSPECV_MWAIT 9)
(UNSPECV_CMPXCHG 10)
(UNSPECV_XCHG 12)
(UNSPECV_LOCK 13)
(UNSPECV_PROLOGUE_USE 14)
(UNSPECV_CLD 15)
(UNSPECV_VZEROALL 16)
(UNSPECV_VZEROUPPER 17)
(UNSPECV_RDTSC 18)
(UNSPECV_RDTSCP 19)
(UNSPECV_RDPMC 20)
])
;; Constants to represent pcomtrue/pcomfalse variants
(define_constants
[(PCOM_FALSE 0)
(PCOM_TRUE 1)
(COM_FALSE_S 2)
(COM_FALSE_P 3)
(COM_TRUE_S 4)
(COM_TRUE_P 5)
])
;; Registers by name.
(define_constants
[(AX_REG 0)
(DX_REG 1)
(CX_REG 2)
(BX_REG 3)
(SI_REG 4)
(DI_REG 5)
(BP_REG 6)
(SP_REG 7)
(ST0_REG 8)
(ST1_REG 9)
(ST2_REG 10)
(ST3_REG 11)
(ST4_REG 12)
(ST5_REG 13)
(ST6_REG 14)
(ST7_REG 15)
(FLAGS_REG 17)
(FPSR_REG 18)
(FPCR_REG 19)
(XMM0_REG 21)
(XMM1_REG 22)
(XMM2_REG 23)
(XMM3_REG 24)
(XMM4_REG 25)
(XMM5_REG 26)
(XMM6_REG 27)
(XMM7_REG 28)
(MM0_REG 29)
(MM1_REG 30)
(MM2_REG 31)
(MM3_REG 32)
(MM4_REG 33)
(MM5_REG 34)
(MM6_REG 35)
(MM7_REG 36)
(R8_REG 37)
(R9_REG 38)
(R10_REG 39)
(R11_REG 40)
(R12_REG 41)
(R13_REG 42)
(XMM8_REG 45)
(XMM9_REG 46)
(XMM10_REG 47)
(XMM11_REG 48)
(XMM12_REG 49)
(XMM13_REG 50)
(XMM14_REG 51)
(XMM15_REG 52)
])
;; Insns whose names begin with "x86_" are emitted by gen_FOO calls
;; from i386.c.
;; In C guard expressions, put expressions which may be compile-time
;; constants first. This allows for better optimization. For
;; example, write "TARGET_64BIT && reload_completed", not
;; "reload_completed && TARGET_64BIT".
;; Processor type.
(define_attr "cpu" "none,pentium,pentiumpro,geode,k6,athlon,k8,core2,atom,
generic64,amdfam10"
(const (symbol_ref "ix86_schedule")))
;; A basic instruction type. Refinements due to arguments to be
;; provided in other attributes.
(define_attr "type"
"other,multi,
alu,alu1,negnot,imov,imovx,lea,
incdec,ishift,ishift1,rotate,rotate1,imul,idiv,
icmp,test,ibr,setcc,icmov,
push,pop,call,callv,leave,
str,bitmanip,
fmov,fop,fsgn,fmul,fdiv,fpspc,fcmov,fcmp,fxch,fistp,fisttp,frndint,
sselog,sselog1,sseiadd,sseiadd1,sseishft,sseimul,
sse,ssemov,sseadd,ssemul,ssecmp,ssecomi,ssecvt,ssecvt1,sseicvt,ssediv,sseins,
ssemuladd,sse4arg,
mmx,mmxmov,mmxadd,mmxmul,mmxcmp,mmxcvt,mmxshft"
(const_string "other"))
;; Main data type used by the insn
(define_attr "mode"
"unknown,none,QI,HI,SI,DI,TI,OI,SF,DF,XF,TF,V8SF,V4DF,V4SF,V2DF,V2SF,V1DF"
(const_string "unknown"))
;; The CPU unit operations uses.
(define_attr "unit" "integer,i387,sse,mmx,unknown"
(cond [(eq_attr "type" "fmov,fop,fsgn,fmul,fdiv,fpspc,fcmov,fcmp,fxch,fistp,fisttp,frndint")
(const_string "i387")
(eq_attr "type" "sselog,sselog1,sseiadd,sseiadd1,sseishft,sseimul,
sse,ssemov,sseadd,ssemul,ssecmp,ssecomi,ssecvt,
ssecvt1,sseicvt,ssediv,sseins,ssemuladd,sse4arg")
(const_string "sse")
(eq_attr "type" "mmx,mmxmov,mmxadd,mmxmul,mmxcmp,mmxcvt,mmxshft")
(const_string "mmx")
(eq_attr "type" "other")
(const_string "unknown")]
(const_string "integer")))
;; The (bounding maximum) length of an instruction immediate.
(define_attr "length_immediate" ""
(cond [(eq_attr "type" "incdec,setcc,icmov,str,lea,other,multi,idiv,leave,
bitmanip")
(const_int 0)
(eq_attr "unit" "i387,sse,mmx")
(const_int 0)
(eq_attr "type" "alu,alu1,negnot,imovx,ishift,rotate,ishift1,rotate1,
imul,icmp,push,pop")
(symbol_ref "ix86_attr_length_immediate_default(insn,1)")
(eq_attr "type" "imov,test")
(symbol_ref "ix86_attr_length_immediate_default(insn,0)")
(eq_attr "type" "call")
(if_then_else (match_operand 0 "constant_call_address_operand" "")
(const_int 4)
(const_int 0))
(eq_attr "type" "callv")
(if_then_else (match_operand 1 "constant_call_address_operand" "")
(const_int 4)
(const_int 0))
;; We don't know the size before shorten_branches. Expect
;; the instruction to fit for better scheduling.
(eq_attr "type" "ibr")
(const_int 1)
]
(symbol_ref "/* Update immediate_length and other attributes! */
gcc_unreachable (),1")))
;; The (bounding maximum) length of an instruction address.
(define_attr "length_address" ""
(cond [(eq_attr "type" "str,other,multi,fxch")
(const_int 0)
(and (eq_attr "type" "call")
(match_operand 0 "constant_call_address_operand" ""))
(const_int 0)
(and (eq_attr "type" "callv")
(match_operand 1 "constant_call_address_operand" ""))
(const_int 0)
]
(symbol_ref "ix86_attr_length_address_default (insn)")))
;; Set when length prefix is used.
(define_attr "prefix_data16" ""
(cond [(eq_attr "type" "ssemuladd,sse4arg,sseiadd1,ssecvt1")
(const_int 0)
(eq_attr "mode" "HI")
(const_int 1)
(and (eq_attr "unit" "sse") (eq_attr "mode" "V2DF,TI"))
(const_int 1)
]
(const_int 0)))
;; Set when string REP prefix is used.
(define_attr "prefix_rep" ""
(cond [(eq_attr "type" "ssemuladd,sse4arg,sseiadd1,ssecvt1")
(const_int 0)
(and (eq_attr "unit" "sse") (eq_attr "mode" "SF,DF"))
(const_int 1)
]
(const_int 0)))
;; Set when 0f opcode prefix is used.
(define_attr "prefix_0f" ""
(if_then_else
(ior (eq_attr "type" "imovx,setcc,icmov,bitmanip")
(eq_attr "unit" "sse,mmx"))
(const_int 1)
(const_int 0)))
;; Set when REX opcode prefix is used.
(define_attr "prefix_rex" ""
(cond [(ne (symbol_ref "!TARGET_64BIT") (const_int 0))
(const_int 0)
(and (eq_attr "mode" "DI")
(and (eq_attr "type" "!push,pop,call,callv,leave,ibr")
(eq_attr "unit" "!mmx")))
(const_int 1)
(and (eq_attr "mode" "QI")
(ne (symbol_ref "x86_extended_QIreg_mentioned_p (insn)")
(const_int 0)))
(const_int 1)
(ne (symbol_ref "x86_extended_reg_mentioned_p (insn)")
(const_int 0))
(const_int 1)
(and (eq_attr "type" "imovx")
(match_operand:QI 1 "ext_QIreg_operand" ""))
(const_int 1)
]
(const_int 0)))
;; There are also additional prefixes in 3DNOW, SSSE3.
;; ssemuladd,sse4arg default to 0f24/0f25 and DREX byte,
;; sseiadd1,ssecvt1 to 0f7a with no DREX byte.
;; 3DNOW has 0f0f prefix, SSSE3 and SSE4_{1,2} 0f38/0f3a.
(define_attr "prefix_extra" ""
(cond [(eq_attr "type" "ssemuladd,sse4arg")
(const_int 2)
(eq_attr "type" "sseiadd1,ssecvt1")
(const_int 1)
]
(const_int 0)))
;; Prefix used: original, VEX or maybe VEX.
(define_attr "prefix" "orig,vex,maybe_vex"
(if_then_else (eq_attr "mode" "OI,V8SF,V4DF")
(const_string "vex")
(const_string "orig")))
;; VEX W bit is used.
(define_attr "prefix_vex_w" "" (const_int 0))
;; The length of VEX prefix
;; Only instructions with 0f prefix can have 2 byte VEX prefix,
;; 0f38/0f3a prefixes can't. In i386.md 0f3[8a] is
;; still prefix_0f 1, with prefix_extra 1.
(define_attr "length_vex" ""
(if_then_else (and (eq_attr "prefix_0f" "1")
(eq_attr "prefix_extra" "0"))
(if_then_else (eq_attr "prefix_vex_w" "1")
(symbol_ref "ix86_attr_length_vex_default (insn, 1, 1)")
(symbol_ref "ix86_attr_length_vex_default (insn, 1, 0)"))
(if_then_else (eq_attr "prefix_vex_w" "1")
(symbol_ref "ix86_attr_length_vex_default (insn, 0, 1)")
(symbol_ref "ix86_attr_length_vex_default (insn, 0, 0)"))))
;; Set when modrm byte is used.
(define_attr "modrm" ""
(cond [(eq_attr "type" "str,leave")
(const_int 0)
(eq_attr "unit" "i387")
(const_int 0)
(and (eq_attr "type" "incdec")
(and (eq (symbol_ref "TARGET_64BIT") (const_int 0))
(ior (match_operand:SI 1 "register_operand" "")
(match_operand:HI 1 "register_operand" ""))))
(const_int 0)
(and (eq_attr "type" "push")
(not (match_operand 1 "memory_operand" "")))
(const_int 0)
(and (eq_attr "type" "pop")
(not (match_operand 0 "memory_operand" "")))
(const_int 0)
(and (eq_attr "type" "imov")
(and (not (eq_attr "mode" "DI"))
(ior (and (match_operand 0 "register_operand" "")
(match_operand 1 "immediate_operand" ""))
(ior (and (match_operand 0 "ax_reg_operand" "")
(match_operand 1 "memory_displacement_only_operand" ""))
(and (match_operand 0 "memory_displacement_only_operand" "")
(match_operand 1 "ax_reg_operand" ""))))))
(const_int 0)
(and (eq_attr "type" "call")
(match_operand 0 "constant_call_address_operand" ""))
(const_int 0)
(and (eq_attr "type" "callv")
(match_operand 1 "constant_call_address_operand" ""))
(const_int 0)
(and (eq_attr "type" "alu,alu1,icmp,test")
(match_operand 0 "ax_reg_operand" ""))
(symbol_ref "(get_attr_length_immediate (insn) <= (get_attr_mode (insn) != MODE_QI))")
]
(const_int 1)))
;; The (bounding maximum) length of an instruction in bytes.
;; ??? fistp and frndint are in fact fldcw/{fistp,frndint}/fldcw sequences.
;; Later we may want to split them and compute proper length as for
;; other insns.
(define_attr "length" ""
(cond [(eq_attr "type" "other,multi,fistp,frndint")
(const_int 16)
(eq_attr "type" "fcmp")
(const_int 4)
(eq_attr "unit" "i387")
(plus (const_int 2)
(plus (attr "prefix_data16")
(attr "length_address")))
(ior (eq_attr "prefix" "vex")
(and (eq_attr "prefix" "maybe_vex")
(ne (symbol_ref "TARGET_AVX") (const_int 0))))
(plus (attr "length_vex")
(plus (attr "length_immediate")
(plus (attr "modrm")
(attr "length_address"))))]
(plus (plus (attr "modrm")
(plus (attr "prefix_0f")
(plus (attr "prefix_rex")
(plus (attr "prefix_extra")
(const_int 1)))))
(plus (attr "prefix_rep")
(plus (attr "prefix_data16")
(plus (attr "length_immediate")
(attr "length_address")))))))
;; The `memory' attribute is `none' if no memory is referenced, `load' or
;; `store' if there is a simple memory reference therein, or `unknown'
;; if the instruction is complex.
(define_attr "memory" "none,load,store,both,unknown"
(cond [(eq_attr "type" "other,multi,str")
(const_string "unknown")
(eq_attr "type" "lea,fcmov,fpspc")
(const_string "none")
(eq_attr "type" "fistp,leave")
(const_string "both")
(eq_attr "type" "frndint")
(const_string "load")
(eq_attr "type" "push")
(if_then_else (match_operand 1 "memory_operand" "")
(const_string "both")
(const_string "store"))
(eq_attr "type" "pop")
(if_then_else (match_operand 0 "memory_operand" "")
(const_string "both")
(const_string "load"))
(eq_attr "type" "setcc")
(if_then_else (match_operand 0 "memory_operand" "")
(const_string "store")
(const_string "none"))
(eq_attr "type" "icmp,test,ssecmp,ssecomi,mmxcmp,fcmp")
(if_then_else (ior (match_operand 0 "memory_operand" "")
(match_operand 1 "memory_operand" ""))
(const_string "load")
(const_string "none"))
(eq_attr "type" "ibr")
(if_then_else (match_operand 0 "memory_operand" "")
(const_string "load")
(const_string "none"))
(eq_attr "type" "call")
(if_then_else (match_operand 0 "constant_call_address_operand" "")
(const_string "none")
(const_string "load"))
(eq_attr "type" "callv")
(if_then_else (match_operand 1 "constant_call_address_operand" "")
(const_string "none")
(const_string "load"))
(and (eq_attr "type" "alu1,negnot,ishift1,sselog1")
(match_operand 1 "memory_operand" ""))
(const_string "both")
(and (match_operand 0 "memory_operand" "")
(match_operand 1 "memory_operand" ""))
(const_string "both")
(match_operand 0 "memory_operand" "")
(const_string "store")
(match_operand 1 "memory_operand" "")
(const_string "load")
(and (eq_attr "type"
"!alu1,negnot,ishift1,
imov,imovx,icmp,test,bitmanip,
fmov,fcmp,fsgn,
sse,ssemov,ssecmp,ssecomi,ssecvt,ssecvt1,sseicvt,sselog1,
sseiadd1,mmx,mmxmov,mmxcmp,mmxcvt")
(match_operand 2 "memory_operand" ""))
(const_string "load")
(and (eq_attr "type" "icmov,ssemuladd,sse4arg")
(match_operand 3 "memory_operand" ""))
(const_string "load")
]
(const_string "none")))
;; Indicates if an instruction has both an immediate and a displacement.
(define_attr "imm_disp" "false,true,unknown"
(cond [(eq_attr "type" "other,multi")
(const_string "unknown")
(and (eq_attr "type" "icmp,test,imov,alu1,ishift1,rotate1")
(and (match_operand 0 "memory_displacement_operand" "")
(match_operand 1 "immediate_operand" "")))
(const_string "true")
(and (eq_attr "type" "alu,ishift,rotate,imul,idiv")
(and (match_operand 0 "memory_displacement_operand" "")
(match_operand 2 "immediate_operand" "")))
(const_string "true")
]
(const_string "false")))
;; Indicates if an FP operation has an integer source.
(define_attr "fp_int_src" "false,true"
(const_string "false"))
;; Defines rounding mode of an FP operation.
(define_attr "i387_cw" "trunc,floor,ceil,mask_pm,uninitialized,any"
(const_string "any"))
;; Define attribute to classify add/sub insns that consumes carry flag (CF)
(define_attr "use_carry" "0,1" (const_string "0"))
;; Define attribute to indicate unaligned ssemov insns
(define_attr "movu" "0,1" (const_string "0"))
;; Describe a user's asm statement.
(define_asm_attributes
[(set_attr "length" "128")
(set_attr "type" "multi")])
;; All integer comparison codes.
(define_code_iterator int_cond [ne eq ge gt le lt geu gtu leu ltu ])
;; All floating-point comparison codes.
(define_code_iterator fp_cond [unordered ordered
uneq unge ungt unle unlt ltgt ])
(define_code_iterator plusminus [plus minus])
(define_code_iterator sat_plusminus [ss_plus us_plus ss_minus us_minus])
;; Base name for define_insn
(define_code_attr plusminus_insn
[(plus "add") (ss_plus "ssadd") (us_plus "usadd")
(minus "sub") (ss_minus "sssub") (us_minus "ussub")])
;; Base name for insn mnemonic.
(define_code_attr plusminus_mnemonic
[(plus "add") (ss_plus "adds") (us_plus "addus")
(minus "sub") (ss_minus "subs") (us_minus "subus")])
;; Mark commutative operators as such in constraints.
(define_code_attr comm [(plus "%") (ss_plus "%") (us_plus "%")
(minus "") (ss_minus "") (us_minus "")])
;; Mapping of signed max and min
(define_code_iterator smaxmin [smax smin])
;; Mapping of unsigned max and min
(define_code_iterator umaxmin [umax umin])
;; Mapping of signed/unsigned max and min
(define_code_iterator maxmin [smax smin umax umin])
;; Base name for integer and FP insn mnemonic
(define_code_attr maxminiprefix [(smax "maxs") (smin "mins")
(umax "maxu") (umin "minu")])
(define_code_attr maxminfprefix [(smax "max") (smin "min")])
;; Mapping of parallel logic operators
(define_code_iterator plogic [and ior xor])
;; Base name for insn mnemonic.
(define_code_attr plogicprefix [(and "and") (ior "or") (xor "xor")])
;; Mapping of abs neg operators
(define_code_iterator absneg [abs neg])
;; Base name for x87 insn mnemonic.
(define_code_attr absnegprefix [(abs "abs") (neg "chs")])
;; Used in signed and unsigned widening multiplications.
(define_code_iterator any_extend [sign_extend zero_extend])
;; Various insn prefixes for widening operations.
(define_code_attr u [(sign_extend "") (zero_extend "u")])
(define_code_attr s [(sign_extend "s") (zero_extend "u")])
;; Instruction prefix for widening operations.
(define_code_attr sgnprefix [(sign_extend "i") (zero_extend "")])
;; All single word integer modes.
(define_mode_iterator SWI [QI HI SI (DI "TARGET_64BIT")])
;; Single word integer modes without QImode.
(define_mode_iterator SWI248 [HI SI (DI "TARGET_64BIT")])
;; Single word integer modes without QImode and HImode.
(define_mode_iterator SWI48 [SI (DI "TARGET_64BIT")])
;; All math-dependant single and double word integer modes.
(define_mode_iterator SDWIM [(QI "TARGET_QIMODE_MATH")
(HI "TARGET_HIMODE_MATH")
SI DI (TI "TARGET_64BIT")])
;; Math-dependant single word integer modes without QImode.
(define_mode_iterator SWIM248 [(HI "TARGET_HIMODE_MATH")
SI (DI "TARGET_64BIT")])
;; Half mode for double word integer modes.
(define_mode_iterator DWIH [(SI "!TARGET_64BIT")
(DI "TARGET_64BIT")])
;; Double word integer modes.
(define_mode_attr DWI [(SI "DI") (DI "TI")])
(define_mode_attr dwi [(SI "di") (DI "ti")])
;; Instruction suffix for integer modes.
(define_mode_attr imodesuffix [(QI "b") (HI "w") (SI "l") (DI "q")])
;; Register class for integer modes.
(define_mode_attr r [(QI "q") (HI "r") (SI "r") (DI "r")])
;; Immediate operand constraint for integer modes.
(define_mode_attr i [(QI "n") (HI "n") (SI "i") (DI "e")])
;; General operand constraint for word modes.
(define_mode_attr g [(SI "g") (DI "rme")])
;; Immediate operand constraint for double integer modes.
(define_mode_attr di [(SI "iF") (DI "e")])
;; General operand predicate for integer modes.
(define_mode_attr general_operand
[(QI "general_operand")
(HI "general_operand")
(SI "general_operand")
(DI "x86_64_general_operand")
(TI "x86_64_general_operand")])
;; SSE and x87 SFmode and DFmode floating point modes
(define_mode_iterator MODEF [SF DF])
;; All x87 floating point modes
(define_mode_iterator X87MODEF [SF DF XF])
;; All integer modes handled by x87 fisttp operator.
(define_mode_iterator X87MODEI [HI SI DI])
;; All integer modes handled by integer x87 operators.
(define_mode_iterator X87MODEI12 [HI SI])
;; All integer modes handled by SSE cvtts?2si* operators.
(define_mode_iterator SSEMODEI24 [SI DI])
;; SSE asm suffix for floating point modes
(define_mode_attr ssemodefsuffix [(SF "s") (DF "d")])
;; SSE vector mode corresponding to a scalar mode
(define_mode_attr ssevecmode
[(QI "V16QI") (HI "V8HI") (SI "V4SI") (DI "V2DI") (SF "V4SF") (DF "V2DF")])
;; Instruction suffix for REX 64bit operators.
(define_mode_attr rex64suffix [(SI "") (DI "{q}")])
;; This mode iterator allows :P to be used for patterns that operate on
;; pointer-sized quantities. Exactly one of the two alternatives will match.
(define_mode_iterator P [(SI "Pmode == SImode") (DI "Pmode == DImode")])
;; Scheduling descriptions
(include "pentium.md")
(include "ppro.md")
(include "k6.md")
(include "athlon.md")
(include "geode.md")
(include "atom.md")
;; Operand and operator predicates and constraints
(include "predicates.md")
(include "constraints.md")
;; Compare and branch/compare and store instructions.
(define_expand "cbranchti4"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:TI 1 "nonimmediate_operand" "")
(match_operand:TI 2 "x86_64_general_operand" "")))
(set (pc) (if_then_else
(match_operator 0 "comparison_operator"
[(reg:CC FLAGS_REG)
(const_int 0)])
(label_ref (match_operand 3 "" ""))
(pc)))]
"TARGET_64BIT"
{
if (MEM_P (operands[1]) && MEM_P (operands[2]))
operands[1] = force_reg (TImode, operands[1]);
ix86_compare_op0 = operands[1];
ix86_compare_op1 = operands[2];
ix86_expand_branch (GET_CODE (operands[0]), operands[3]);
DONE;
})
(define_expand "cbranchdi4"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:DI 1 "nonimmediate_operand" "")
(match_operand:DI 2 "x86_64_general_operand" "")))
(set (pc) (if_then_else
(match_operator 0 "comparison_operator"
[(reg:CC FLAGS_REG)
(const_int 0)])
(label_ref (match_operand 3 "" ""))
(pc)))]
""
{
if (MEM_P (operands[1]) && MEM_P (operands[2]))
operands[1] = force_reg (DImode, operands[1]);
ix86_compare_op0 = operands[1];
ix86_compare_op1 = operands[2];
ix86_expand_branch (GET_CODE (operands[0]), operands[3]);
DONE;
})
(define_expand "cstoredi4"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:DI 2 "nonimmediate_operand" "")
(match_operand:DI 3 "x86_64_general_operand" "")))
(set (match_operand:QI 0 "register_operand" "")
(match_operator 1 "comparison_operator"
[(reg:CC FLAGS_REG)
(const_int 0)]))]
"TARGET_64BIT"
{
if (MEM_P (operands[2]) && MEM_P (operands[3]))
operands[2] = force_reg (DImode, operands[2]);
ix86_compare_op0 = operands[2];
ix86_compare_op1 = operands[3];
ix86_expand_setcc (GET_CODE (operands[1]), operands[0]);
DONE;
})
(define_expand "cbranchsi4"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:SI 1 "cmpsi_operand" "")
(match_operand:SI 2 "general_operand" "")))
(set (pc) (if_then_else
(match_operator 0 "comparison_operator"
[(reg:CC FLAGS_REG)
(const_int 0)])
(label_ref (match_operand 3 "" ""))
(pc)))]
""
{
if (MEM_P (operands[1]) && MEM_P (operands[2]))
operands[1] = force_reg (SImode, operands[1]);
ix86_compare_op0 = operands[1];
ix86_compare_op1 = operands[2];
ix86_expand_branch (GET_CODE (operands[0]), operands[3]);
DONE;
})
(define_expand "cstoresi4"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:SI 2 "cmpsi_operand" "")
(match_operand:SI 3 "general_operand" "")))
(set (match_operand:QI 0 "register_operand" "")
(match_operator 1 "comparison_operator"
[(reg:CC FLAGS_REG)
(const_int 0)]))]
""
{
if (MEM_P (operands[2]) && MEM_P (operands[3]))
operands[2] = force_reg (SImode, operands[2]);
ix86_compare_op0 = operands[2];
ix86_compare_op1 = operands[3];
ix86_expand_setcc (GET_CODE (operands[1]), operands[0]);
DONE;
})
(define_expand "cbranchhi4"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:HI 1 "nonimmediate_operand" "")
(match_operand:HI 2 "general_operand" "")))
(set (pc) (if_then_else
(match_operator 0 "comparison_operator"
[(reg:CC FLAGS_REG)
(const_int 0)])
(label_ref (match_operand 3 "" ""))
(pc)))]
""
{
if (MEM_P (operands[1]) && MEM_P (operands[2]))
operands[1] = force_reg (HImode, operands[1]);
ix86_compare_op0 = operands[1];
ix86_compare_op1 = operands[2];
ix86_expand_branch (GET_CODE (operands[0]), operands[3]);
DONE;
})
(define_expand "cstorehi4"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:HI 2 "nonimmediate_operand" "")
(match_operand:HI 3 "general_operand" "")))
(set (match_operand:QI 0 "register_operand" "")
(match_operator 1 "comparison_operator"
[(reg:CC FLAGS_REG)
(const_int 0)]))]
""
{
if (MEM_P (operands[2]) && MEM_P (operands[3]))
operands[2] = force_reg (HImode, operands[2]);
ix86_compare_op0 = operands[2];
ix86_compare_op1 = operands[3];
ix86_expand_setcc (GET_CODE (operands[1]), operands[0]);
DONE;
})
(define_expand "cbranchqi4"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:QI 1 "nonimmediate_operand" "")
(match_operand:QI 2 "general_operand" "")))
(set (pc) (if_then_else
(match_operator 0 "comparison_operator"
[(reg:CC FLAGS_REG)
(const_int 0)])
(label_ref (match_operand 3 "" ""))
(pc)))]
""
{
if (MEM_P (operands[1]) && MEM_P (operands[2]))
operands[1] = force_reg (QImode, operands[1]);
ix86_compare_op0 = operands[1];
ix86_compare_op1 = operands[2];
ix86_expand_branch (GET_CODE (operands[0]), operands[3]);
DONE;
})
(define_expand "cstoreqi4"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:QI 2 "nonimmediate_operand" "")
(match_operand:QI 3 "general_operand" "")))
(set (match_operand:QI 0 "register_operand" "")
(match_operator 1 "comparison_operator"
[(reg:CC FLAGS_REG)
(const_int 0)]))]
""
{
if (MEM_P (operands[2]) && MEM_P (operands[3]))
operands[2] = force_reg (QImode, operands[2]);
ix86_compare_op0 = operands[2];
ix86_compare_op1 = operands[3];
ix86_expand_setcc (GET_CODE (operands[1]), operands[0]);
DONE;
})
(define_insn "cmpdi_ccno_1_rex64"
[(set (reg FLAGS_REG)
(compare (match_operand:DI 0 "nonimmediate_operand" "r,?mr")
(match_operand:DI 1 "const0_operand" "")))]
"TARGET_64BIT && ix86_match_ccmode (insn, CCNOmode)"
"@
test{q}\t%0, %0
cmp{q}\t{%1, %0|%0, %1}"
[(set_attr "type" "test,icmp")
(set_attr "length_immediate" "0,1")
(set_attr "mode" "DI")])
(define_insn "*cmpdi_minus_1_rex64"
[(set (reg FLAGS_REG)
(compare (minus:DI (match_operand:DI 0 "nonimmediate_operand" "rm,r")
(match_operand:DI 1 "x86_64_general_operand" "re,mr"))
(const_int 0)))]
"TARGET_64BIT && ix86_match_ccmode (insn, CCGOCmode)"
"cmp{q}\t{%1, %0|%0, %1}"
[(set_attr "type" "icmp")
(set_attr "mode" "DI")])
(define_expand "cmpdi_1_rex64"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:DI 0 "nonimmediate_operand" "")
(match_operand:DI 1 "general_operand" "")))]
"TARGET_64BIT"
"")
(define_insn "cmpdi_1_insn_rex64"
[(set (reg FLAGS_REG)
(compare (match_operand:DI 0 "nonimmediate_operand" "mr,r")
(match_operand:DI 1 "x86_64_general_operand" "re,mr")))]
"TARGET_64BIT && ix86_match_ccmode (insn, CCmode)"
"cmp{q}\t{%1, %0|%0, %1}"
[(set_attr "type" "icmp")
(set_attr "mode" "DI")])
(define_insn "*cmpsi_ccno_1"
[(set (reg FLAGS_REG)
(compare (match_operand:SI 0 "nonimmediate_operand" "r,?mr")
(match_operand:SI 1 "const0_operand" "")))]
"ix86_match_ccmode (insn, CCNOmode)"
"@
test{l}\t%0, %0
cmp{l}\t{%1, %0|%0, %1}"
[(set_attr "type" "test,icmp")
(set_attr "length_immediate" "0,1")
(set_attr "mode" "SI")])
(define_insn "*cmpsi_minus_1"
[(set (reg FLAGS_REG)
(compare (minus:SI (match_operand:SI 0 "nonimmediate_operand" "rm,r")
(match_operand:SI 1 "general_operand" "ri,mr"))
(const_int 0)))]
"ix86_match_ccmode (insn, CCGOCmode)"
"cmp{l}\t{%1, %0|%0, %1}"
[(set_attr "type" "icmp")
(set_attr "mode" "SI")])
(define_expand "cmpsi_1"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:SI 0 "nonimmediate_operand" "")
(match_operand:SI 1 "general_operand" "")))]
""
"")
(define_insn "*cmpsi_1_insn"
[(set (reg FLAGS_REG)
(compare (match_operand:SI 0 "nonimmediate_operand" "rm,r")
(match_operand:SI 1 "general_operand" "ri,mr")))]
"!(MEM_P (operands[0]) && MEM_P (operands[1]))
&& ix86_match_ccmode (insn, CCmode)"
"cmp{l}\t{%1, %0|%0, %1}"
[(set_attr "type" "icmp")
(set_attr "mode" "SI")])
(define_insn "*cmphi_ccno_1"
[(set (reg FLAGS_REG)
(compare (match_operand:HI 0 "nonimmediate_operand" "r,?mr")
(match_operand:HI 1 "const0_operand" "")))]
"ix86_match_ccmode (insn, CCNOmode)"
"@
test{w}\t%0, %0
cmp{w}\t{%1, %0|%0, %1}"
[(set_attr "type" "test,icmp")
(set_attr "length_immediate" "0,1")
(set_attr "mode" "HI")])
(define_insn "*cmphi_minus_1"
[(set (reg FLAGS_REG)
(compare (minus:HI (match_operand:HI 0 "nonimmediate_operand" "rm,r")
(match_operand:HI 1 "general_operand" "rn,mr"))
(const_int 0)))]
"ix86_match_ccmode (insn, CCGOCmode)"
"cmp{w}\t{%1, %0|%0, %1}"
[(set_attr "type" "icmp")
(set_attr "mode" "HI")])
(define_insn "*cmphi_1"
[(set (reg FLAGS_REG)
(compare (match_operand:HI 0 "nonimmediate_operand" "rm,r")
(match_operand:HI 1 "general_operand" "rn,mr")))]
"!(MEM_P (operands[0]) && MEM_P (operands[1]))
&& ix86_match_ccmode (insn, CCmode)"
"cmp{w}\t{%1, %0|%0, %1}"
[(set_attr "type" "icmp")
(set_attr "mode" "HI")])
(define_insn "*cmpqi_ccno_1"
[(set (reg FLAGS_REG)
(compare (match_operand:QI 0 "nonimmediate_operand" "q,?mq")
(match_operand:QI 1 "const0_operand" "")))]
"ix86_match_ccmode (insn, CCNOmode)"
"@
test{b}\t%0, %0
cmp{b}\t{$0, %0|%0, 0}"
[(set_attr "type" "test,icmp")
(set_attr "length_immediate" "0,1")
(set_attr "mode" "QI")])
(define_insn "*cmpqi_1"
[(set (reg FLAGS_REG)
(compare (match_operand:QI 0 "nonimmediate_operand" "qm,q")
(match_operand:QI 1 "general_operand" "qn,mq")))]
"!(MEM_P (operands[0]) && MEM_P (operands[1]))
&& ix86_match_ccmode (insn, CCmode)"
"cmp{b}\t{%1, %0|%0, %1}"
[(set_attr "type" "icmp")
(set_attr "mode" "QI")])
(define_insn "*cmpqi_minus_1"
[(set (reg FLAGS_REG)
(compare (minus:QI (match_operand:QI 0 "nonimmediate_operand" "qm,q")
(match_operand:QI 1 "general_operand" "qn,mq"))
(const_int 0)))]
"ix86_match_ccmode (insn, CCGOCmode)"
"cmp{b}\t{%1, %0|%0, %1}"
[(set_attr "type" "icmp")
(set_attr "mode" "QI")])
(define_insn "*cmpqi_ext_1"
[(set (reg FLAGS_REG)
(compare
(match_operand:QI 0 "general_operand" "Qm")
(subreg:QI
(zero_extract:SI
(match_operand 1 "ext_register_operand" "Q")
(const_int 8)
(const_int 8)) 0)))]
"!TARGET_64BIT && ix86_match_ccmode (insn, CCmode)"
"cmp{b}\t{%h1, %0|%0, %h1}"
[(set_attr "type" "icmp")
(set_attr "mode" "QI")])
(define_insn "*cmpqi_ext_1_rex64"
[(set (reg FLAGS_REG)
(compare
(match_operand:QI 0 "register_operand" "Q")
(subreg:QI
(zero_extract:SI
(match_operand 1 "ext_register_operand" "Q")
(const_int 8)
(const_int 8)) 0)))]
"TARGET_64BIT && ix86_match_ccmode (insn, CCmode)"
"cmp{b}\t{%h1, %0|%0, %h1}"
[(set_attr "type" "icmp")
(set_attr "mode" "QI")])
(define_insn "*cmpqi_ext_2"
[(set (reg FLAGS_REG)
(compare
(subreg:QI
(zero_extract:SI
(match_operand 0 "ext_register_operand" "Q")
(const_int 8)
(const_int 8)) 0)
(match_operand:QI 1 "const0_operand" "")))]
"ix86_match_ccmode (insn, CCNOmode)"
"test{b}\t%h0, %h0"
[(set_attr "type" "test")
(set_attr "length_immediate" "0")
(set_attr "mode" "QI")])
(define_expand "cmpqi_ext_3"
[(set (reg:CC FLAGS_REG)
(compare:CC
(subreg:QI
(zero_extract:SI
(match_operand 0 "ext_register_operand" "")
(const_int 8)
(const_int 8)) 0)
(match_operand:QI 1 "general_operand" "")))]
""
"")
(define_insn "cmpqi_ext_3_insn"
[(set (reg FLAGS_REG)
(compare
(subreg:QI
(zero_extract:SI
(match_operand 0 "ext_register_operand" "Q")
(const_int 8)
(const_int 8)) 0)
(match_operand:QI 1 "general_operand" "Qmn")))]
"!TARGET_64BIT && ix86_match_ccmode (insn, CCmode)"
"cmp{b}\t{%1, %h0|%h0, %1}"
[(set_attr "type" "icmp")
(set_attr "modrm" "1")
(set_attr "mode" "QI")])
(define_insn "cmpqi_ext_3_insn_rex64"
[(set (reg FLAGS_REG)
(compare
(subreg:QI
(zero_extract:SI
(match_operand 0 "ext_register_operand" "Q")
(const_int 8)
(const_int 8)) 0)
(match_operand:QI 1 "nonmemory_operand" "Qn")))]
"TARGET_64BIT && ix86_match_ccmode (insn, CCmode)"
"cmp{b}\t{%1, %h0|%h0, %1}"
[(set_attr "type" "icmp")
(set_attr "modrm" "1")
(set_attr "mode" "QI")])
(define_insn "*cmpqi_ext_4"
[(set (reg FLAGS_REG)
(compare
(subreg:QI
(zero_extract:SI
(match_operand 0 "ext_register_operand" "Q")
(const_int 8)
(const_int 8)) 0)
(subreg:QI
(zero_extract:SI
(match_operand 1 "ext_register_operand" "Q")
(const_int 8)
(const_int 8)) 0)))]
"ix86_match_ccmode (insn, CCmode)"
"cmp{b}\t{%h1, %h0|%h0, %h1}"
[(set_attr "type" "icmp")
(set_attr "mode" "QI")])
;; These implement float point compares.
;; %%% See if we can get away with VOIDmode operands on the actual insns,
;; which would allow mix and match FP modes on the compares. Which is what
;; the old patterns did, but with many more of them.
(define_expand "cbranchxf4"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:XF 1 "nonmemory_operand" "")
(match_operand:XF 2 "nonmemory_operand" "")))
(set (pc) (if_then_else
(match_operator 0 "ix86_fp_comparison_operator"
[(reg:CC FLAGS_REG)
(const_int 0)])
(label_ref (match_operand 3 "" ""))
(pc)))]
"TARGET_80387"
{
ix86_compare_op0 = operands[1];
ix86_compare_op1 = operands[2];
ix86_expand_branch (GET_CODE (operands[0]), operands[3]);
DONE;
})
(define_expand "cstorexf4"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:XF 2 "nonmemory_operand" "")
(match_operand:XF 3 "nonmemory_operand" "")))
(set (match_operand:QI 0 "register_operand" "")
(match_operator 1 "ix86_fp_comparison_operator"
[(reg:CC FLAGS_REG)
(const_int 0)]))]
"TARGET_80387"
{
ix86_compare_op0 = operands[2];
ix86_compare_op1 = operands[3];
ix86_expand_setcc (GET_CODE (operands[1]), operands[0]);
DONE;
})
(define_expand "cbranch4"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:MODEF 1 "cmp_fp_expander_operand" "")
(match_operand:MODEF 2 "cmp_fp_expander_operand" "")))
(set (pc) (if_then_else
(match_operator 0 "ix86_fp_comparison_operator"
[(reg:CC FLAGS_REG)
(const_int 0)])
(label_ref (match_operand 3 "" ""))
(pc)))]
"TARGET_80387 || (SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH)"
{
ix86_compare_op0 = operands[1];
ix86_compare_op1 = operands[2];
ix86_expand_branch (GET_CODE (operands[0]), operands[3]);
DONE;
})
(define_expand "cstore4"
[(set (reg:CC FLAGS_REG)
(compare:CC (match_operand:MODEF 2 "cmp_fp_expander_operand" "")
(match_operand:MODEF 3 "cmp_fp_expander_operand" "")))
(set (match_operand:QI 0 "register_operand" "")
(match_operator 1 "ix86_fp_comparison_operator"
[(reg:CC FLAGS_REG)
(const_int 0)]))]
"TARGET_80387 || (SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH)"
{
ix86_compare_op0 = operands[2];
ix86_compare_op1 = operands[3];
ix86_expand_setcc (GET_CODE (operands[1]), operands[0]);
DONE;
})
(define_expand "cbranchcc4"
[(set (pc) (if_then_else
(match_operator 0 "comparison_operator"
[(match_operand 1 "flags_reg_operand" "")
(match_operand 2 "const0_operand" "")])
(label_ref (match_operand 3 "" ""))
(pc)))]
""
{
ix86_compare_op0 = operands[1];
ix86_compare_op1 = operands[2];
ix86_expand_branch (GET_CODE (operands[0]), operands[3]);
DONE;
})
(define_expand "cstorecc4"
[(set (match_operand:QI 0 "register_operand" "")
(match_operator 1 "comparison_operator"
[(match_operand 2 "flags_reg_operand" "")
(match_operand 3 "const0_operand" "")]))]
""
{
ix86_compare_op0 = operands[2];
ix86_compare_op1 = operands[3];
ix86_expand_setcc (GET_CODE (operands[1]), operands[0]);
DONE;
})
;; FP compares, step 1:
;; Set the FP condition codes.
;;
;; CCFPmode compare with exceptions
;; CCFPUmode compare with no exceptions
;; We may not use "#" to split and emit these, since the REG_DEAD notes
;; used to manage the reg stack popping would not be preserved.
(define_insn "*cmpfp_0"
[(set (match_operand:HI 0 "register_operand" "=a")
(unspec:HI
[(compare:CCFP
(match_operand 1 "register_operand" "f")
(match_operand 2 "const0_operand" ""))]
UNSPEC_FNSTSW))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& GET_MODE (operands[1]) == GET_MODE (operands[2])"
"* return output_fp_compare (insn, operands, 0, 0);"
[(set_attr "type" "multi")
(set_attr "unit" "i387")
(set (attr "mode")
(cond [(match_operand:SF 1 "" "")
(const_string "SF")
(match_operand:DF 1 "" "")
(const_string "DF")
]
(const_string "XF")))])
(define_insn_and_split "*cmpfp_0_cc"
[(set (reg:CCFP FLAGS_REG)
(compare:CCFP
(match_operand 1 "register_operand" "f")
(match_operand 2 "const0_operand" "")))
(clobber (match_operand:HI 0 "register_operand" "=a"))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& TARGET_SAHF && !TARGET_CMOVE
&& GET_MODE (operands[1]) == GET_MODE (operands[2])"
"#"
"&& reload_completed"
[(set (match_dup 0)
(unspec:HI
[(compare:CCFP (match_dup 1)(match_dup 2))]
UNSPEC_FNSTSW))
(set (reg:CC FLAGS_REG)
(unspec:CC [(match_dup 0)] UNSPEC_SAHF))]
""
[(set_attr "type" "multi")
(set_attr "unit" "i387")
(set (attr "mode")
(cond [(match_operand:SF 1 "" "")
(const_string "SF")
(match_operand:DF 1 "" "")
(const_string "DF")
]
(const_string "XF")))])
(define_insn "*cmpfp_xf"
[(set (match_operand:HI 0 "register_operand" "=a")
(unspec:HI
[(compare:CCFP
(match_operand:XF 1 "register_operand" "f")
(match_operand:XF 2 "register_operand" "f"))]
UNSPEC_FNSTSW))]
"TARGET_80387"
"* return output_fp_compare (insn, operands, 0, 0);"
[(set_attr "type" "multi")
(set_attr "unit" "i387")
(set_attr "mode" "XF")])
(define_insn_and_split "*cmpfp_xf_cc"
[(set (reg:CCFP FLAGS_REG)
(compare:CCFP
(match_operand:XF 1 "register_operand" "f")
(match_operand:XF 2 "register_operand" "f")))
(clobber (match_operand:HI 0 "register_operand" "=a"))]
"TARGET_80387
&& TARGET_SAHF && !TARGET_CMOVE"
"#"
"&& reload_completed"
[(set (match_dup 0)
(unspec:HI
[(compare:CCFP (match_dup 1)(match_dup 2))]
UNSPEC_FNSTSW))
(set (reg:CC FLAGS_REG)
(unspec:CC [(match_dup 0)] UNSPEC_SAHF))]
""
[(set_attr "type" "multi")
(set_attr "unit" "i387")
(set_attr "mode" "XF")])
(define_insn "*cmpfp_"
[(set (match_operand:HI 0 "register_operand" "=a")
(unspec:HI
[(compare:CCFP
(match_operand:MODEF 1 "register_operand" "f")
(match_operand:MODEF 2 "nonimmediate_operand" "fm"))]
UNSPEC_FNSTSW))]
"TARGET_80387"
"* return output_fp_compare (insn, operands, 0, 0);"
[(set_attr "type" "multi")
(set_attr "unit" "i387")
(set_attr "mode" "")])
(define_insn_and_split "*cmpfp__cc"
[(set (reg:CCFP FLAGS_REG)
(compare:CCFP
(match_operand:MODEF 1 "register_operand" "f")
(match_operand:MODEF 2 "nonimmediate_operand" "fm")))
(clobber (match_operand:HI 0 "register_operand" "=a"))]
"TARGET_80387
&& TARGET_SAHF && !TARGET_CMOVE"
"#"
"&& reload_completed"
[(set (match_dup 0)
(unspec:HI
[(compare:CCFP (match_dup 1)(match_dup 2))]
UNSPEC_FNSTSW))
(set (reg:CC FLAGS_REG)
(unspec:CC [(match_dup 0)] UNSPEC_SAHF))]
""
[(set_attr "type" "multi")
(set_attr "unit" "i387")
(set_attr "mode" "")])
(define_insn "*cmpfp_u"
[(set (match_operand:HI 0 "register_operand" "=a")
(unspec:HI
[(compare:CCFPU
(match_operand 1 "register_operand" "f")
(match_operand 2 "register_operand" "f"))]
UNSPEC_FNSTSW))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& GET_MODE (operands[1]) == GET_MODE (operands[2])"
"* return output_fp_compare (insn, operands, 0, 1);"
[(set_attr "type" "multi")
(set_attr "unit" "i387")
(set (attr "mode")
(cond [(match_operand:SF 1 "" "")
(const_string "SF")
(match_operand:DF 1 "" "")
(const_string "DF")
]
(const_string "XF")))])
(define_insn_and_split "*cmpfp_u_cc"
[(set (reg:CCFPU FLAGS_REG)
(compare:CCFPU
(match_operand 1 "register_operand" "f")
(match_operand 2 "register_operand" "f")))
(clobber (match_operand:HI 0 "register_operand" "=a"))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& TARGET_SAHF && !TARGET_CMOVE
&& GET_MODE (operands[1]) == GET_MODE (operands[2])"
"#"
"&& reload_completed"
[(set (match_dup 0)
(unspec:HI
[(compare:CCFPU (match_dup 1)(match_dup 2))]
UNSPEC_FNSTSW))
(set (reg:CC FLAGS_REG)
(unspec:CC [(match_dup 0)] UNSPEC_SAHF))]
""
[(set_attr "type" "multi")
(set_attr "unit" "i387")
(set (attr "mode")
(cond [(match_operand:SF 1 "" "")
(const_string "SF")
(match_operand:DF 1 "" "")
(const_string "DF")
]
(const_string "XF")))])
(define_insn "*cmpfp_"
[(set (match_operand:HI 0 "register_operand" "=a")
(unspec:HI
[(compare:CCFP
(match_operand 1 "register_operand" "f")
(match_operator 3 "float_operator"
[(match_operand:X87MODEI12 2 "memory_operand" "m")]))]
UNSPEC_FNSTSW))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& (TARGET_USE_MODE_FIOP || optimize_function_for_size_p (cfun))
&& (GET_MODE (operands [3]) == GET_MODE (operands[1]))"
"* return output_fp_compare (insn, operands, 0, 0);"
[(set_attr "type" "multi")
(set_attr "unit" "i387")
(set_attr "fp_int_src" "true")
(set_attr "mode" "")])
(define_insn_and_split "*cmpfp__cc"
[(set (reg:CCFP FLAGS_REG)
(compare:CCFP
(match_operand 1 "register_operand" "f")
(match_operator 3 "float_operator"
[(match_operand:X87MODEI12 2 "memory_operand" "m")])))
(clobber (match_operand:HI 0 "register_operand" "=a"))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& TARGET_SAHF && !TARGET_CMOVE
&& (TARGET_USE_MODE_FIOP || optimize_function_for_size_p (cfun))
&& (GET_MODE (operands [3]) == GET_MODE (operands[1]))"
"#"
"&& reload_completed"
[(set (match_dup 0)
(unspec:HI
[(compare:CCFP
(match_dup 1)
(match_op_dup 3 [(match_dup 2)]))]
UNSPEC_FNSTSW))
(set (reg:CC FLAGS_REG)
(unspec:CC [(match_dup 0)] UNSPEC_SAHF))]
""
[(set_attr "type" "multi")
(set_attr "unit" "i387")
(set_attr "fp_int_src" "true")
(set_attr "mode" "")])
;; FP compares, step 2
;; Move the fpsw to ax.
(define_insn "x86_fnstsw_1"
[(set (match_operand:HI 0 "register_operand" "=a")
(unspec:HI [(reg:CCFP FPSR_REG)] UNSPEC_FNSTSW))]
"TARGET_80387"
"fnstsw\t%0"
[(set (attr "length") (symbol_ref "ix86_attr_length_address_default (insn) + 2"))
(set_attr "mode" "SI")
(set_attr "unit" "i387")])
;; FP compares, step 3
;; Get ax into flags, general case.
(define_insn "x86_sahf_1"
[(set (reg:CC FLAGS_REG)
(unspec:CC [(match_operand:HI 0 "register_operand" "a")]
UNSPEC_SAHF))]
"TARGET_SAHF"
{
#ifdef HAVE_AS_IX86_SAHF
return "sahf";
#else
return ASM_BYTE "0x9e";
#endif
}
[(set_attr "length" "1")
(set_attr "athlon_decode" "vector")
(set_attr "amdfam10_decode" "direct")
(set_attr "mode" "SI")])
;; Pentium Pro can do steps 1 through 3 in one go.
;; comi*, ucomi*, fcomi*, ficomi*,fucomi* (i387 instructions set condition codes)
(define_insn "*cmpfp_i_mixed"
[(set (reg:CCFP FLAGS_REG)
(compare:CCFP (match_operand 0 "register_operand" "f,x")
(match_operand 1 "nonimmediate_operand" "f,xm")))]
"TARGET_MIX_SSE_I387
&& SSE_FLOAT_MODE_P (GET_MODE (operands[0]))
&& GET_MODE (operands[0]) == GET_MODE (operands[1])"
"* return output_fp_compare (insn, operands, 1, 0);"
[(set_attr "type" "fcmp,ssecomi")
(set_attr "prefix" "orig,maybe_vex")
(set (attr "mode")
(if_then_else (match_operand:SF 1 "" "")
(const_string "SF")
(const_string "DF")))
(set (attr "prefix_rep")
(if_then_else (eq_attr "type" "ssecomi")
(const_string "0")
(const_string "*")))
(set (attr "prefix_data16")
(cond [(eq_attr "type" "fcmp")
(const_string "*")
(eq_attr "mode" "DF")
(const_string "1")
]
(const_string "0")))
(set_attr "athlon_decode" "vector")
(set_attr "amdfam10_decode" "direct")])
(define_insn "*cmpfp_i_sse"
[(set (reg:CCFP FLAGS_REG)
(compare:CCFP (match_operand 0 "register_operand" "x")
(match_operand 1 "nonimmediate_operand" "xm")))]
"TARGET_SSE_MATH
&& SSE_FLOAT_MODE_P (GET_MODE (operands[0]))
&& GET_MODE (operands[0]) == GET_MODE (operands[1])"
"* return output_fp_compare (insn, operands, 1, 0);"
[(set_attr "type" "ssecomi")
(set_attr "prefix" "maybe_vex")
(set (attr "mode")
(if_then_else (match_operand:SF 1 "" "")
(const_string "SF")
(const_string "DF")))
(set_attr "prefix_rep" "0")
(set (attr "prefix_data16")
(if_then_else (eq_attr "mode" "DF")
(const_string "1")
(const_string "0")))
(set_attr "athlon_decode" "vector")
(set_attr "amdfam10_decode" "direct")])
(define_insn "*cmpfp_i_i387"
[(set (reg:CCFP FLAGS_REG)
(compare:CCFP (match_operand 0 "register_operand" "f")
(match_operand 1 "register_operand" "f")))]
"X87_FLOAT_MODE_P (GET_MODE (operands[0]))
&& TARGET_CMOVE
&& !(SSE_FLOAT_MODE_P (GET_MODE (operands[0])) && TARGET_SSE_MATH)
&& GET_MODE (operands[0]) == GET_MODE (operands[1])"
"* return output_fp_compare (insn, operands, 1, 0);"
[(set_attr "type" "fcmp")
(set (attr "mode")
(cond [(match_operand:SF 1 "" "")
(const_string "SF")
(match_operand:DF 1 "" "")
(const_string "DF")
]
(const_string "XF")))
(set_attr "athlon_decode" "vector")
(set_attr "amdfam10_decode" "direct")])
(define_insn "*cmpfp_iu_mixed"
[(set (reg:CCFPU FLAGS_REG)
(compare:CCFPU (match_operand 0 "register_operand" "f,x")
(match_operand 1 "nonimmediate_operand" "f,xm")))]
"TARGET_MIX_SSE_I387
&& SSE_FLOAT_MODE_P (GET_MODE (operands[0]))
&& GET_MODE (operands[0]) == GET_MODE (operands[1])"
"* return output_fp_compare (insn, operands, 1, 1);"
[(set_attr "type" "fcmp,ssecomi")
(set_attr "prefix" "orig,maybe_vex")
(set (attr "mode")
(if_then_else (match_operand:SF 1 "" "")
(const_string "SF")
(const_string "DF")))
(set (attr "prefix_rep")
(if_then_else (eq_attr "type" "ssecomi")
(const_string "0")
(const_string "*")))
(set (attr "prefix_data16")
(cond [(eq_attr "type" "fcmp")
(const_string "*")
(eq_attr "mode" "DF")
(const_string "1")
]
(const_string "0")))
(set_attr "athlon_decode" "vector")
(set_attr "amdfam10_decode" "direct")])
(define_insn "*cmpfp_iu_sse"
[(set (reg:CCFPU FLAGS_REG)
(compare:CCFPU (match_operand 0 "register_operand" "x")
(match_operand 1 "nonimmediate_operand" "xm")))]
"TARGET_SSE_MATH
&& SSE_FLOAT_MODE_P (GET_MODE (operands[0]))
&& GET_MODE (operands[0]) == GET_MODE (operands[1])"
"* return output_fp_compare (insn, operands, 1, 1);"
[(set_attr "type" "ssecomi")
(set_attr "prefix" "maybe_vex")
(set (attr "mode")
(if_then_else (match_operand:SF 1 "" "")
(const_string "SF")
(const_string "DF")))
(set_attr "prefix_rep" "0")
(set (attr "prefix_data16")
(if_then_else (eq_attr "mode" "DF")
(const_string "1")
(const_string "0")))
(set_attr "athlon_decode" "vector")
(set_attr "amdfam10_decode" "direct")])
(define_insn "*cmpfp_iu_387"
[(set (reg:CCFPU FLAGS_REG)
(compare:CCFPU (match_operand 0 "register_operand" "f")
(match_operand 1 "register_operand" "f")))]
"X87_FLOAT_MODE_P (GET_MODE (operands[0]))
&& TARGET_CMOVE
&& !(SSE_FLOAT_MODE_P (GET_MODE (operands[0])) && TARGET_SSE_MATH)
&& GET_MODE (operands[0]) == GET_MODE (operands[1])"
"* return output_fp_compare (insn, operands, 1, 1);"
[(set_attr "type" "fcmp")
(set (attr "mode")
(cond [(match_operand:SF 1 "" "")
(const_string "SF")
(match_operand:DF 1 "" "")
(const_string "DF")
]
(const_string "XF")))
(set_attr "athlon_decode" "vector")
(set_attr "amdfam10_decode" "direct")])
;; Move instructions.
;; General case of fullword move.
(define_expand "movsi"
[(set (match_operand:SI 0 "nonimmediate_operand" "")
(match_operand:SI 1 "general_operand" ""))]
""
"ix86_expand_move (SImode, operands); DONE;")
;; Push/pop instructions. They are separate since autoinc/dec is not a
;; general_operand.
;;
;; %%% We don't use a post-inc memory reference because x86 is not a
;; general AUTO_INC_DEC host, which impacts how it is treated in flow.
;; Changing this impacts compiler performance on other non-AUTO_INC_DEC
;; targets without our curiosities, and it is just as easy to represent
;; this differently.
(define_insn "*pushsi2"
[(set (match_operand:SI 0 "push_operand" "=<")
(match_operand:SI 1 "general_no_elim_operand" "ri*m"))]
"!TARGET_64BIT"
"push{l}\t%1"
[(set_attr "type" "push")
(set_attr "mode" "SI")])
;; For 64BIT abi we always round up to 8 bytes.
(define_insn "*pushsi2_rex64"
[(set (match_operand:SI 0 "push_operand" "=X")
(match_operand:SI 1 "nonmemory_no_elim_operand" "ri"))]
"TARGET_64BIT"
"push{q}\t%q1"
[(set_attr "type" "push")
(set_attr "mode" "SI")])
(define_insn "*pushsi2_prologue"
[(set (match_operand:SI 0 "push_operand" "=<")
(match_operand:SI 1 "general_no_elim_operand" "ri*m"))
(clobber (mem:BLK (scratch)))]
"!TARGET_64BIT"
"push{l}\t%1"
[(set_attr "type" "push")
(set_attr "mode" "SI")])
(define_insn "*popsi1_epilogue"
[(set (match_operand:SI 0 "nonimmediate_operand" "=r*m")
(mem:SI (reg:SI SP_REG)))
(set (reg:SI SP_REG)
(plus:SI (reg:SI SP_REG) (const_int 4)))
(clobber (mem:BLK (scratch)))]
"!TARGET_64BIT"
"pop{l}\t%0"
[(set_attr "type" "pop")
(set_attr "mode" "SI")])
(define_insn "popsi1"
[(set (match_operand:SI 0 "nonimmediate_operand" "=r*m")
(mem:SI (reg:SI SP_REG)))
(set (reg:SI SP_REG)
(plus:SI (reg:SI SP_REG) (const_int 4)))]
"!TARGET_64BIT"
"pop{l}\t%0"
[(set_attr "type" "pop")
(set_attr "mode" "SI")])
(define_insn "*movsi_xor"
[(set (match_operand:SI 0 "register_operand" "=r")
(match_operand:SI 1 "const0_operand" ""))
(clobber (reg:CC FLAGS_REG))]
"reload_completed"
"xor{l}\t%0, %0"
[(set_attr "type" "alu1")
(set_attr "mode" "SI")
(set_attr "length_immediate" "0")])
(define_insn "*movsi_or"
[(set (match_operand:SI 0 "register_operand" "=r")
(match_operand:SI 1 "immediate_operand" "i"))
(clobber (reg:CC FLAGS_REG))]
"reload_completed
&& operands[1] == constm1_rtx"
{
operands[1] = constm1_rtx;
return "or{l}\t{%1, %0|%0, %1}";
}
[(set_attr "type" "alu1")
(set_attr "mode" "SI")
(set_attr "length_immediate" "1")])
(define_insn "*movsi_1"
[(set (match_operand:SI 0 "nonimmediate_operand"
"=r,m ,*y,*y,?rm,?*y,*x,*x,?r ,m ,?*Yi,*x")
(match_operand:SI 1 "general_operand"
"g ,ri,C ,*y,*y ,rm ,C ,*x,*Yi,*x,r ,m "))]
"!(MEM_P (operands[0]) && MEM_P (operands[1]))"
{
switch (get_attr_type (insn))
{
case TYPE_SSELOG1:
if (get_attr_mode (insn) == MODE_TI)
return "%vpxor\t%0, %d0";
return "%vxorps\t%0, %d0";
case TYPE_SSEMOV:
switch (get_attr_mode (insn))
{
case MODE_TI:
return "%vmovdqa\t{%1, %0|%0, %1}";
case MODE_V4SF:
return "%vmovaps\t{%1, %0|%0, %1}";
case MODE_SI:
return "%vmovd\t{%1, %0|%0, %1}";
case MODE_SF:
return "%vmovss\t{%1, %0|%0, %1}";
default:
gcc_unreachable ();
}
case TYPE_MMX:
return "pxor\t%0, %0";
case TYPE_MMXMOV:
if (get_attr_mode (insn) == MODE_DI)
return "movq\t{%1, %0|%0, %1}";
return "movd\t{%1, %0|%0, %1}";
case TYPE_LEA:
return "lea{l}\t{%1, %0|%0, %1}";
default:
gcc_assert (!flag_pic || LEGITIMATE_PIC_OPERAND_P (operands[1]));
return "mov{l}\t{%1, %0|%0, %1}";
}
}
[(set (attr "type")
(cond [(eq_attr "alternative" "2")
(const_string "mmx")
(eq_attr "alternative" "3,4,5")
(const_string "mmxmov")
(eq_attr "alternative" "6")
(const_string "sselog1")
(eq_attr "alternative" "7,8,9,10,11")
(const_string "ssemov")
(match_operand:DI 1 "pic_32bit_operand" "")
(const_string "lea")
]
(const_string "imov")))
(set (attr "prefix")
(if_then_else (eq_attr "alternative" "0,1,2,3,4,5")
(const_string "orig")
(const_string "maybe_vex")))
(set (attr "prefix_data16")
(if_then_else (and (eq_attr "type" "ssemov") (eq_attr "mode" "SI"))
(const_string "1")
(const_string "*")))
(set (attr "mode")
(cond [(eq_attr "alternative" "2,3")
(const_string "DI")
(eq_attr "alternative" "6,7")
(if_then_else
(eq (symbol_ref "TARGET_SSE2") (const_int 0))
(const_string "V4SF")
(const_string "TI"))
(and (eq_attr "alternative" "8,9,10,11")
(eq (symbol_ref "TARGET_SSE2") (const_int 0)))
(const_string "SF")
]
(const_string "SI")))])
;; Stores and loads of ax to arbitrary constant address.
;; We fake an second form of instruction to force reload to load address
;; into register when rax is not available
(define_insn "*movabssi_1_rex64"
[(set (mem:SI (match_operand:DI 0 "x86_64_movabs_operand" "i,r"))
(match_operand:SI 1 "nonmemory_operand" "a,er"))]
"TARGET_64BIT && ix86_check_movabs (insn, 0)"
"@
movabs{l}\t{%1, %P0|%P0, %1}
mov{l}\t{%1, %a0|%a0, %1}"
[(set_attr "type" "imov")
(set_attr "modrm" "0,*")
(set_attr "length_address" "8,0")
(set_attr "length_immediate" "0,*")
(set_attr "memory" "store")
(set_attr "mode" "SI")])
(define_insn "*movabssi_2_rex64"
[(set (match_operand:SI 0 "register_operand" "=a,r")
(mem:SI (match_operand:DI 1 "x86_64_movabs_operand" "i,r")))]
"TARGET_64BIT && ix86_check_movabs (insn, 1)"
"@
movabs{l}\t{%P1, %0|%0, %P1}
mov{l}\t{%a1, %0|%0, %a1}"
[(set_attr "type" "imov")
(set_attr "modrm" "0,*")
(set_attr "length_address" "8,0")
(set_attr "length_immediate" "0")
(set_attr "memory" "load")
(set_attr "mode" "SI")])
(define_insn "*swapsi"
[(set (match_operand:SI 0 "register_operand" "+r")
(match_operand:SI 1 "register_operand" "+r"))
(set (match_dup 1)
(match_dup 0))]
""
"xchg{l}\t%1, %0"
[(set_attr "type" "imov")
(set_attr "mode" "SI")
(set_attr "pent_pair" "np")
(set_attr "athlon_decode" "vector")
(set_attr "amdfam10_decode" "double")])
(define_expand "movhi"
[(set (match_operand:HI 0 "nonimmediate_operand" "")
(match_operand:HI 1 "general_operand" ""))]
""
"ix86_expand_move (HImode, operands); DONE;")
(define_insn "*pushhi2"
[(set (match_operand:HI 0 "push_operand" "=X")
(match_operand:HI 1 "nonmemory_no_elim_operand" "rn"))]
"!TARGET_64BIT"
"push{l}\t%k1"
[(set_attr "type" "push")
(set_attr "mode" "SI")])
;; For 64BIT abi we always round up to 8 bytes.
(define_insn "*pushhi2_rex64"
[(set (match_operand:HI 0 "push_operand" "=X")
(match_operand:HI 1 "nonmemory_no_elim_operand" "rn"))]
"TARGET_64BIT"
"push{q}\t%q1"
[(set_attr "type" "push")
(set_attr "mode" "DI")])
(define_insn "*movhi_1"
[(set (match_operand:HI 0 "nonimmediate_operand" "=r,r,r,m")
(match_operand:HI 1 "general_operand" "r,rn,rm,rn"))]
"!(MEM_P (operands[0]) && MEM_P (operands[1]))"
{
switch (get_attr_type (insn))
{
case TYPE_IMOVX:
/* movzwl is faster than movw on p2 due to partial word stalls,
though not as fast as an aligned movl. */
return "movz{wl|x}\t{%1, %k0|%k0, %1}";
default:
if (get_attr_mode (insn) == MODE_SI)
return "mov{l}\t{%k1, %k0|%k0, %k1}";
else
return "mov{w}\t{%1, %0|%0, %1}";
}
}
[(set (attr "type")
(cond [(ne (symbol_ref "optimize_function_for_size_p (cfun)") (const_int 0))
(const_string "imov")
(and (eq_attr "alternative" "0")
(ior (eq (symbol_ref "TARGET_PARTIAL_REG_STALL")
(const_int 0))
(eq (symbol_ref "TARGET_HIMODE_MATH")
(const_int 0))))
(const_string "imov")
(and (eq_attr "alternative" "1,2")
(match_operand:HI 1 "aligned_operand" ""))
(const_string "imov")
(and (ne (symbol_ref "TARGET_MOVX")
(const_int 0))
(eq_attr "alternative" "0,2"))
(const_string "imovx")
]
(const_string "imov")))
(set (attr "mode")
(cond [(eq_attr "type" "imovx")
(const_string "SI")
(and (eq_attr "alternative" "1,2")
(match_operand:HI 1 "aligned_operand" ""))
(const_string "SI")
(and (eq_attr "alternative" "0")
(ior (eq (symbol_ref "TARGET_PARTIAL_REG_STALL")
(const_int 0))
(eq (symbol_ref "TARGET_HIMODE_MATH")
(const_int 0))))
(const_string "SI")
]
(const_string "HI")))])
;; Stores and loads of ax to arbitrary constant address.
;; We fake an second form of instruction to force reload to load address
;; into register when rax is not available
(define_insn "*movabshi_1_rex64"
[(set (mem:HI (match_operand:DI 0 "x86_64_movabs_operand" "i,r"))
(match_operand:HI 1 "nonmemory_operand" "a,er"))]
"TARGET_64BIT && ix86_check_movabs (insn, 0)"
"@
movabs{w}\t{%1, %P0|%P0, %1}
mov{w}\t{%1, %a0|%a0, %1}"
[(set_attr "type" "imov")
(set_attr "modrm" "0,*")
(set_attr "length_address" "8,0")
(set_attr "length_immediate" "0,*")
(set_attr "memory" "store")
(set_attr "mode" "HI")])
(define_insn "*movabshi_2_rex64"
[(set (match_operand:HI 0 "register_operand" "=a,r")
(mem:HI (match_operand:DI 1 "x86_64_movabs_operand" "i,r")))]
"TARGET_64BIT && ix86_check_movabs (insn, 1)"
"@
movabs{w}\t{%P1, %0|%0, %P1}
mov{w}\t{%a1, %0|%0, %a1}"
[(set_attr "type" "imov")
(set_attr "modrm" "0,*")
(set_attr "length_address" "8,0")
(set_attr "length_immediate" "0")
(set_attr "memory" "load")
(set_attr "mode" "HI")])
(define_insn "*swaphi_1"
[(set (match_operand:HI 0 "register_operand" "+r")
(match_operand:HI 1 "register_operand" "+r"))
(set (match_dup 1)
(match_dup 0))]
"!TARGET_PARTIAL_REG_STALL || optimize_function_for_size_p (cfun)"
"xchg{l}\t%k1, %k0"
[(set_attr "type" "imov")
(set_attr "mode" "SI")
(set_attr "pent_pair" "np")
(set_attr "athlon_decode" "vector")
(set_attr "amdfam10_decode" "double")])
;; Not added amdfam10_decode since TARGET_PARTIAL_REG_STALL is disabled for AMDFAM10
(define_insn "*swaphi_2"
[(set (match_operand:HI 0 "register_operand" "+r")
(match_operand:HI 1 "register_operand" "+r"))
(set (match_dup 1)
(match_dup 0))]
"TARGET_PARTIAL_REG_STALL"
"xchg{w}\t%1, %0"
[(set_attr "type" "imov")
(set_attr "mode" "HI")
(set_attr "pent_pair" "np")
(set_attr "athlon_decode" "vector")])
(define_expand "movstricthi"
[(set (strict_low_part (match_operand:HI 0 "nonimmediate_operand" ""))
(match_operand:HI 1 "general_operand" ""))]
""
{
if (TARGET_PARTIAL_REG_STALL && optimize_function_for_speed_p (cfun))
FAIL;
/* Don't generate memory->memory moves, go through a register */
if (MEM_P (operands[0]) && MEM_P (operands[1]))
operands[1] = force_reg (HImode, operands[1]);
})
(define_insn "*movstricthi_1"
[(set (strict_low_part (match_operand:HI 0 "nonimmediate_operand" "+rm,r"))
(match_operand:HI 1 "general_operand" "rn,m"))]
"(! TARGET_PARTIAL_REG_STALL || optimize_function_for_size_p (cfun))
&& !(MEM_P (operands[0]) && MEM_P (operands[1]))"
"mov{w}\t{%1, %0|%0, %1}"
[(set_attr "type" "imov")
(set_attr "mode" "HI")])
(define_insn "*movstricthi_xor"
[(set (strict_low_part (match_operand:HI 0 "register_operand" "+r"))
(match_operand:HI 1 "const0_operand" ""))
(clobber (reg:CC FLAGS_REG))]
"reload_completed"
"xor{w}\t%0, %0"
[(set_attr "type" "alu1")
(set_attr "mode" "HI")
(set_attr "length_immediate" "0")])
(define_expand "movqi"
[(set (match_operand:QI 0 "nonimmediate_operand" "")
(match_operand:QI 1 "general_operand" ""))]
""
"ix86_expand_move (QImode, operands); DONE;")
;; emit_push_insn when it calls move_by_pieces requires an insn to
;; "push a byte". But actually we use pushl, which has the effect
;; of rounding the amount pushed up to a word.
(define_insn "*pushqi2"
[(set (match_operand:QI 0 "push_operand" "=X")
(match_operand:QI 1 "nonmemory_no_elim_operand" "rn"))]
"!TARGET_64BIT"
"push{l}\t%k1"
[(set_attr "type" "push")
(set_attr "mode" "SI")])
;; For 64BIT abi we always round up to 8 bytes.
(define_insn "*pushqi2_rex64"
[(set (match_operand:QI 0 "push_operand" "=X")
(match_operand:QI 1 "nonmemory_no_elim_operand" "qn"))]
"TARGET_64BIT"
"push{q}\t%q1"
[(set_attr "type" "push")
(set_attr "mode" "DI")])
;; Situation is quite tricky about when to choose full sized (SImode) move
;; over QImode moves. For Q_REG -> Q_REG move we use full size only for
;; partial register dependency machines (such as AMD Athlon), where QImode
;; moves issue extra dependency and for partial register stalls machines
;; that don't use QImode patterns (and QImode move cause stall on the next
;; instruction).
;;
;; For loads of Q_REG to NONQ_REG we use full sized moves except for partial
;; register stall machines with, where we use QImode instructions, since
;; partial register stall can be caused there. Then we use movzx.
(define_insn "*movqi_1"
[(set (match_operand:QI 0 "nonimmediate_operand" "=q,q ,q ,r,r ,?r,m")
(match_operand:QI 1 "general_operand" " q,qn,qm,q,rn,qm,qn"))]
"!(MEM_P (operands[0]) && MEM_P (operands[1]))"
{
switch (get_attr_type (insn))
{
case TYPE_IMOVX:
gcc_assert (ANY_QI_REG_P (operands[1]) || MEM_P (operands[1]));
return "movz{bl|x}\t{%1, %k0|%k0, %1}";
default:
if (get_attr_mode (insn) == MODE_SI)
return "mov{l}\t{%k1, %k0|%k0, %k1}";
else
return "mov{b}\t{%1, %0|%0, %1}";
}
}
[(set (attr "type")
(cond [(and (eq_attr "alternative" "5")
(not (match_operand:QI 1 "aligned_operand" "")))
(const_string "imovx")
(ne (symbol_ref "optimize_function_for_size_p (cfun)") (const_int 0))
(const_string "imov")
(and (eq_attr "alternative" "3")
(ior (eq (symbol_ref "TARGET_PARTIAL_REG_STALL")
(const_int 0))
(eq (symbol_ref "TARGET_QIMODE_MATH")
(const_int 0))))
(const_string "imov")
(eq_attr "alternative" "3,5")
(const_string "imovx")
(and (ne (symbol_ref "TARGET_MOVX")
(const_int 0))
(eq_attr "alternative" "2"))
(const_string "imovx")
]
(const_string "imov")))
(set (attr "mode")
(cond [(eq_attr "alternative" "3,4,5")
(const_string "SI")
(eq_attr "alternative" "6")
(const_string "QI")
(eq_attr "type" "imovx")
(const_string "SI")
(and (eq_attr "type" "imov")
(and (eq_attr "alternative" "0,1")
(and (ne (symbol_ref "TARGET_PARTIAL_REG_DEPENDENCY")
(const_int 0))
(and (eq (symbol_ref "optimize_function_for_size_p (cfun)")
(const_int 0))
(eq (symbol_ref "TARGET_PARTIAL_REG_STALL")
(const_int 0))))))
(const_string "SI")
;; Avoid partial register stalls when not using QImode arithmetic
(and (eq_attr "type" "imov")
(and (eq_attr "alternative" "0,1")
(and (ne (symbol_ref "TARGET_PARTIAL_REG_STALL")
(const_int 0))
(eq (symbol_ref "TARGET_QIMODE_MATH")
(const_int 0)))))
(const_string "SI")
]
(const_string "QI")))])
(define_insn "*swapqi_1"
[(set (match_operand:QI 0 "register_operand" "+r")
(match_operand:QI 1 "register_operand" "+r"))
(set (match_dup 1)
(match_dup 0))]
"!TARGET_PARTIAL_REG_STALL || optimize_function_for_size_p (cfun)"
"xchg{l}\t%k1, %k0"
[(set_attr "type" "imov")
(set_attr "mode" "SI")
(set_attr "pent_pair" "np")
(set_attr "athlon_decode" "vector")
(set_attr "amdfam10_decode" "vector")])
;; Not added amdfam10_decode since TARGET_PARTIAL_REG_STALL is disabled for AMDFAM10
(define_insn "*swapqi_2"
[(set (match_operand:QI 0 "register_operand" "+q")
(match_operand:QI 1 "register_operand" "+q"))
(set (match_dup 1)
(match_dup 0))]
"TARGET_PARTIAL_REG_STALL"
"xchg{b}\t%1, %0"
[(set_attr "type" "imov")
(set_attr "mode" "QI")
(set_attr "pent_pair" "np")
(set_attr "athlon_decode" "vector")])
(define_expand "movstrictqi"
[(set (strict_low_part (match_operand:QI 0 "nonimmediate_operand" ""))
(match_operand:QI 1 "general_operand" ""))]
""
{
if (TARGET_PARTIAL_REG_STALL && optimize_function_for_speed_p (cfun))
FAIL;
/* Don't generate memory->memory moves, go through a register. */
if (MEM_P (operands[0]) && MEM_P (operands[1]))
operands[1] = force_reg (QImode, operands[1]);
})
(define_insn "*movstrictqi_1"
[(set (strict_low_part (match_operand:QI 0 "nonimmediate_operand" "+qm,q"))
(match_operand:QI 1 "general_operand" "*qn,m"))]
"(! TARGET_PARTIAL_REG_STALL || optimize_function_for_size_p (cfun))
&& !(MEM_P (operands[0]) && MEM_P (operands[1]))"
"mov{b}\t{%1, %0|%0, %1}"
[(set_attr "type" "imov")
(set_attr "mode" "QI")])
(define_insn "*movstrictqi_xor"
[(set (strict_low_part (match_operand:QI 0 "q_regs_operand" "+q"))
(match_operand:QI 1 "const0_operand" ""))
(clobber (reg:CC FLAGS_REG))]
"reload_completed"
"xor{b}\t%0, %0"
[(set_attr "type" "alu1")
(set_attr "mode" "QI")
(set_attr "length_immediate" "0")])
(define_insn "*movsi_extv_1"
[(set (match_operand:SI 0 "register_operand" "=R")
(sign_extract:SI (match_operand 1 "ext_register_operand" "Q")
(const_int 8)
(const_int 8)))]
""
"movs{bl|x}\t{%h1, %0|%0, %h1}"
[(set_attr "type" "imovx")
(set_attr "mode" "SI")])
(define_insn "*movhi_extv_1"
[(set (match_operand:HI 0 "register_operand" "=R")
(sign_extract:HI (match_operand 1 "ext_register_operand" "Q")
(const_int 8)
(const_int 8)))]
""
"movs{bl|x}\t{%h1, %k0|%k0, %h1}"
[(set_attr "type" "imovx")
(set_attr "mode" "SI")])
(define_insn "*movqi_extv_1"
[(set (match_operand:QI 0 "nonimmediate_operand" "=Qm,?r")
(sign_extract:QI (match_operand 1 "ext_register_operand" "Q,Q")
(const_int 8)
(const_int 8)))]
"!TARGET_64BIT"
{
switch (get_attr_type (insn))
{
case TYPE_IMOVX:
return "movs{bl|x}\t{%h1, %k0|%k0, %h1}";
default:
return "mov{b}\t{%h1, %0|%0, %h1}";
}
}
[(set (attr "type")
(if_then_else (and (match_operand:QI 0 "register_operand" "")
(ior (not (match_operand:QI 0 "q_regs_operand" ""))
(ne (symbol_ref "TARGET_MOVX")
(const_int 0))))
(const_string "imovx")
(const_string "imov")))
(set (attr "mode")
(if_then_else (eq_attr "type" "imovx")
(const_string "SI")
(const_string "QI")))])
(define_insn "*movqi_extv_1_rex64"
[(set (match_operand:QI 0 "register_operand" "=Q,?R")
(sign_extract:QI (match_operand 1 "ext_register_operand" "Q,Q")
(const_int 8)
(const_int 8)))]
"TARGET_64BIT"
{
switch (get_attr_type (insn))
{
case TYPE_IMOVX:
return "movs{bl|x}\t{%h1, %k0|%k0, %h1}";
default:
return "mov{b}\t{%h1, %0|%0, %h1}";
}
}
[(set (attr "type")
(if_then_else (and (match_operand:QI 0 "register_operand" "")
(ior (not (match_operand:QI 0 "q_regs_operand" ""))
(ne (symbol_ref "TARGET_MOVX")
(const_int 0))))
(const_string "imovx")
(const_string "imov")))
(set (attr "mode")
(if_then_else (eq_attr "type" "imovx")
(const_string "SI")
(const_string "QI")))])
;; Stores and loads of ax to arbitrary constant address.
;; We fake an second form of instruction to force reload to load address
;; into register when rax is not available
(define_insn "*movabsqi_1_rex64"
[(set (mem:QI (match_operand:DI 0 "x86_64_movabs_operand" "i,r"))
(match_operand:QI 1 "nonmemory_operand" "a,er"))]
"TARGET_64BIT && ix86_check_movabs (insn, 0)"
"@
movabs{b}\t{%1, %P0|%P0, %1}
mov{b}\t{%1, %a0|%a0, %1}"
[(set_attr "type" "imov")
(set_attr "modrm" "0,*")
(set_attr "length_address" "8,0")
(set_attr "length_immediate" "0,*")
(set_attr "memory" "store")
(set_attr "mode" "QI")])
(define_insn "*movabsqi_2_rex64"
[(set (match_operand:QI 0 "register_operand" "=a,r")
(mem:QI (match_operand:DI 1 "x86_64_movabs_operand" "i,r")))]
"TARGET_64BIT && ix86_check_movabs (insn, 1)"
"@
movabs{b}\t{%P1, %0|%0, %P1}
mov{b}\t{%a1, %0|%0, %a1}"
[(set_attr "type" "imov")
(set_attr "modrm" "0,*")
(set_attr "length_address" "8,0")
(set_attr "length_immediate" "0")
(set_attr "memory" "load")
(set_attr "mode" "QI")])
(define_insn "*movdi_extzv_1"
[(set (match_operand:DI 0 "register_operand" "=R")
(zero_extract:DI (match_operand 1 "ext_register_operand" "Q")
(const_int 8)
(const_int 8)))]
"TARGET_64BIT"
"movz{bl|x}\t{%h1, %k0|%k0, %h1}"
[(set_attr "type" "imovx")
(set_attr "mode" "SI")])
(define_insn "*movsi_extzv_1"
[(set (match_operand:SI 0 "register_operand" "=R")
(zero_extract:SI (match_operand 1 "ext_register_operand" "Q")
(const_int 8)
(const_int 8)))]
""
"movz{bl|x}\t{%h1, %0|%0, %h1}"
[(set_attr "type" "imovx")
(set_attr "mode" "SI")])
(define_insn "*movqi_extzv_2"
[(set (match_operand:QI 0 "nonimmediate_operand" "=Qm,?R")
(subreg:QI (zero_extract:SI (match_operand 1 "ext_register_operand" "Q,Q")
(const_int 8)
(const_int 8)) 0))]
"!TARGET_64BIT"
{
switch (get_attr_type (insn))
{
case TYPE_IMOVX:
return "movz{bl|x}\t{%h1, %k0|%k0, %h1}";
default:
return "mov{b}\t{%h1, %0|%0, %h1}";
}
}
[(set (attr "type")
(if_then_else (and (match_operand:QI 0 "register_operand" "")
(ior (not (match_operand:QI 0 "q_regs_operand" ""))
(ne (symbol_ref "TARGET_MOVX")
(const_int 0))))
(const_string "imovx")
(const_string "imov")))
(set (attr "mode")
(if_then_else (eq_attr "type" "imovx")
(const_string "SI")
(const_string "QI")))])
(define_insn "*movqi_extzv_2_rex64"
[(set (match_operand:QI 0 "register_operand" "=Q,?R")
(subreg:QI (zero_extract:SI (match_operand 1 "ext_register_operand" "Q,Q")
(const_int 8)
(const_int 8)) 0))]
"TARGET_64BIT"
{
switch (get_attr_type (insn))
{
case TYPE_IMOVX:
return "movz{bl|x}\t{%h1, %k0|%k0, %h1}";
default:
return "mov{b}\t{%h1, %0|%0, %h1}";
}
}
[(set (attr "type")
(if_then_else (ior (not (match_operand:QI 0 "q_regs_operand" ""))
(ne (symbol_ref "TARGET_MOVX")
(const_int 0)))
(const_string "imovx")
(const_string "imov")))
(set (attr "mode")
(if_then_else (eq_attr "type" "imovx")
(const_string "SI")
(const_string "QI")))])
(define_insn "movsi_insv_1"
[(set (zero_extract:SI (match_operand 0 "ext_register_operand" "+Q")
(const_int 8)
(const_int 8))
(match_operand:SI 1 "general_operand" "Qmn"))]
"!TARGET_64BIT"
"mov{b}\t{%b1, %h0|%h0, %b1}"
[(set_attr "type" "imov")
(set_attr "mode" "QI")])
(define_insn "*movsi_insv_1_rex64"
[(set (zero_extract:SI (match_operand 0 "ext_register_operand" "+Q")
(const_int 8)
(const_int 8))
(match_operand:SI 1 "nonmemory_operand" "Qn"))]
"TARGET_64BIT"
"mov{b}\t{%b1, %h0|%h0, %b1}"
[(set_attr "type" "imov")
(set_attr "mode" "QI")])
(define_insn "movdi_insv_1_rex64"
[(set (zero_extract:DI (match_operand 0 "ext_register_operand" "+Q")
(const_int 8)
(const_int 8))
(match_operand:DI 1 "nonmemory_operand" "Qn"))]
"TARGET_64BIT"
"mov{b}\t{%b1, %h0|%h0, %b1}"
[(set_attr "type" "imov")
(set_attr "mode" "QI")])
(define_insn "*movqi_insv_2"
[(set (zero_extract:SI (match_operand 0 "ext_register_operand" "+Q")
(const_int 8)
(const_int 8))
(lshiftrt:SI (match_operand:SI 1 "register_operand" "Q")
(const_int 8)))]
""
"mov{b}\t{%h1, %h0|%h0, %h1}"
[(set_attr "type" "imov")
(set_attr "mode" "QI")])
(define_expand "movdi"
[(set (match_operand:DI 0 "nonimmediate_operand" "")
(match_operand:DI 1 "general_operand" ""))]
""
"ix86_expand_move (DImode, operands); DONE;")
(define_insn "*pushdi"
[(set (match_operand:DI 0 "push_operand" "=<")
(match_operand:DI 1 "general_no_elim_operand" "riF*m"))]
"!TARGET_64BIT"
"#")
(define_insn "*pushdi2_rex64"
[(set (match_operand:DI 0 "push_operand" "=<,!<")
(match_operand:DI 1 "general_no_elim_operand" "re*m,n"))]
"TARGET_64BIT"
"@
push{q}\t%1
#"
[(set_attr "type" "push,multi")
(set_attr "mode" "DI")])
;; Convert impossible pushes of immediate to existing instructions.
;; First try to get scratch register and go through it. In case this
;; fails, push sign extended lower part first and then overwrite
;; upper part by 32bit move.
(define_peephole2
[(match_scratch:DI 2 "r")
(set (match_operand:DI 0 "push_operand" "")
(match_operand:DI 1 "immediate_operand" ""))]
"TARGET_64BIT && !symbolic_operand (operands[1], DImode)
&& !x86_64_immediate_operand (operands[1], DImode)"
[(set (match_dup 2) (match_dup 1))
(set (match_dup 0) (match_dup 2))]
"")
;; We need to define this as both peepholer and splitter for case
;; peephole2 pass is not run.
;; "&& 1" is needed to keep it from matching the previous pattern.
(define_peephole2
[(set (match_operand:DI 0 "push_operand" "")
(match_operand:DI 1 "immediate_operand" ""))]
"TARGET_64BIT && !symbolic_operand (operands[1], DImode)
&& !x86_64_immediate_operand (operands[1], DImode) && 1"
[(set (match_dup 0) (match_dup 1))
(set (match_dup 2) (match_dup 3))]
"split_di (&operands[1], 1, &operands[2], &operands[3]);
operands[1] = gen_lowpart (DImode, operands[2]);
operands[2] = gen_rtx_MEM (SImode, gen_rtx_PLUS (DImode, stack_pointer_rtx,
GEN_INT (4)));
")
(define_split
[(set (match_operand:DI 0 "push_operand" "")
(match_operand:DI 1 "immediate_operand" ""))]
"TARGET_64BIT && ((optimize > 0 && flag_peephole2)
? epilogue_completed : reload_completed)
&& !symbolic_operand (operands[1], DImode)
&& !x86_64_immediate_operand (operands[1], DImode)"
[(set (match_dup 0) (match_dup 1))
(set (match_dup 2) (match_dup 3))]
"split_di (&operands[1], 1, &operands[2], &operands[3]);
operands[1] = gen_lowpart (DImode, operands[2]);
operands[2] = gen_rtx_MEM (SImode, gen_rtx_PLUS (DImode, stack_pointer_rtx,
GEN_INT (4)));
")
(define_insn "*pushdi2_prologue_rex64"
[(set (match_operand:DI 0 "push_operand" "=<")
(match_operand:DI 1 "general_no_elim_operand" "re*m"))
(clobber (mem:BLK (scratch)))]
"TARGET_64BIT"
"push{q}\t%1"
[(set_attr "type" "push")
(set_attr "mode" "DI")])
(define_insn "*popdi1_epilogue_rex64"
[(set (match_operand:DI 0 "nonimmediate_operand" "=r*m")
(mem:DI (reg:DI SP_REG)))
(set (reg:DI SP_REG)
(plus:DI (reg:DI SP_REG) (const_int 8)))
(clobber (mem:BLK (scratch)))]
"TARGET_64BIT"
"pop{q}\t%0"
[(set_attr "type" "pop")
(set_attr "mode" "DI")])
(define_insn "popdi1"
[(set (match_operand:DI 0 "nonimmediate_operand" "=r*m")
(mem:DI (reg:DI SP_REG)))
(set (reg:DI SP_REG)
(plus:DI (reg:DI SP_REG) (const_int 8)))]
"TARGET_64BIT"
"pop{q}\t%0"
[(set_attr "type" "pop")
(set_attr "mode" "DI")])
(define_insn "*movdi_xor_rex64"
[(set (match_operand:DI 0 "register_operand" "=r")
(match_operand:DI 1 "const0_operand" ""))
(clobber (reg:CC FLAGS_REG))]
"TARGET_64BIT
&& reload_completed"
"xor{l}\t%k0, %k0";
[(set_attr "type" "alu1")
(set_attr "mode" "SI")
(set_attr "length_immediate" "0")])
(define_insn "*movdi_or_rex64"
[(set (match_operand:DI 0 "register_operand" "=r")
(match_operand:DI 1 "const_int_operand" "i"))
(clobber (reg:CC FLAGS_REG))]
"TARGET_64BIT
&& reload_completed
&& operands[1] == constm1_rtx"
{
operands[1] = constm1_rtx;
return "or{q}\t{%1, %0|%0, %1}";
}
[(set_attr "type" "alu1")
(set_attr "mode" "DI")
(set_attr "length_immediate" "1")])
(define_insn "*movdi_2"
[(set (match_operand:DI 0 "nonimmediate_operand"
"=r ,o ,*y,m*y,*y,*Y2,m ,*Y2,*Y2,*x,m ,*x,*x")
(match_operand:DI 1 "general_operand"
"riFo,riF,C ,*y ,m ,C ,*Y2,*Y2,m ,C ,*x,*x,m "))]
"!TARGET_64BIT && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
"@
#
#
pxor\t%0, %0
movq\t{%1, %0|%0, %1}
movq\t{%1, %0|%0, %1}
%vpxor\t%0, %d0
%vmovq\t{%1, %0|%0, %1}
%vmovdqa\t{%1, %0|%0, %1}
%vmovq\t{%1, %0|%0, %1}
xorps\t%0, %0
movlps\t{%1, %0|%0, %1}
movaps\t{%1, %0|%0, %1}
movlps\t{%1, %0|%0, %1}"
[(set_attr "type" "*,*,mmx,mmxmov,mmxmov,sselog1,ssemov,ssemov,ssemov,sselog1,ssemov,ssemov,ssemov")
(set (attr "prefix")
(if_then_else (eq_attr "alternative" "5,6,7,8")
(const_string "vex")
(const_string "orig")))
(set_attr "mode" "DI,DI,DI,DI,DI,TI,DI,TI,DI,V4SF,V2SF,V4SF,V2SF")])
(define_split
[(set (match_operand:DI 0 "push_operand" "")
(match_operand:DI 1 "general_operand" ""))]
"!TARGET_64BIT && reload_completed
&& (! MMX_REG_P (operands[1]) && !SSE_REG_P (operands[1]))"
[(const_int 0)]
"ix86_split_long_move (operands); DONE;")
;; %%% This multiword shite has got to go.
(define_split
[(set (match_operand:DI 0 "nonimmediate_operand" "")
(match_operand:DI 1 "general_operand" ""))]
"!TARGET_64BIT && reload_completed
&& (!MMX_REG_P (operands[0]) && !SSE_REG_P (operands[0]))
&& (!MMX_REG_P (operands[1]) && !SSE_REG_P (operands[1]))"
[(const_int 0)]
"ix86_split_long_move (operands); DONE;")
(define_insn "*movdi_1_rex64"
[(set (match_operand:DI 0 "nonimmediate_operand"
"=r,r ,r,m ,!m,*y,*y,?r ,m ,?*Ym,?*y,*x,*x,?r ,m,?*Yi,*x,?*x,?*Ym")
(match_operand:DI 1 "general_operand"
"Z ,rem,i,re,n ,C ,*y,*Ym,*y,r ,m ,C ,*x,*Yi,*x,r ,m ,*Ym,*x"))]
"TARGET_64BIT && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
{
switch (get_attr_type (insn))
{
case TYPE_SSECVT:
if (SSE_REG_P (operands[0]))
return "movq2dq\t{%1, %0|%0, %1}";
else
return "movdq2q\t{%1, %0|%0, %1}";
case TYPE_SSEMOV:
if (TARGET_AVX)
{
if (get_attr_mode (insn) == MODE_TI)
return "vmovdqa\t{%1, %0|%0, %1}";
else
return "vmovq\t{%1, %0|%0, %1}";
}
if (get_attr_mode (insn) == MODE_TI)
return "movdqa\t{%1, %0|%0, %1}";
/* FALLTHRU */
case TYPE_MMXMOV:
/* Moves from and into integer register is done using movd
opcode with REX prefix. */
if (GENERAL_REG_P (operands[0]) || GENERAL_REG_P (operands[1]))
return "movd\t{%1, %0|%0, %1}";
return "movq\t{%1, %0|%0, %1}";
case TYPE_SSELOG1:
return "%vpxor\t%0, %d0";
case TYPE_MMX:
return "pxor\t%0, %0";
case TYPE_MULTI:
return "#";
case TYPE_LEA:
return "lea{q}\t{%a1, %0|%0, %a1}";
default:
gcc_assert (!flag_pic || LEGITIMATE_PIC_OPERAND_P (operands[1]));
if (get_attr_mode (insn) == MODE_SI)
return "mov{l}\t{%k1, %k0|%k0, %k1}";
else if (which_alternative == 2)
return "movabs{q}\t{%1, %0|%0, %1}";
else
return "mov{q}\t{%1, %0|%0, %1}";
}
}
[(set (attr "type")
(cond [(eq_attr "alternative" "5")
(const_string "mmx")
(eq_attr "alternative" "6,7,8,9,10")
(const_string "mmxmov")
(eq_attr "alternative" "11")
(const_string "sselog1")
(eq_attr "alternative" "12,13,14,15,16")
(const_string "ssemov")
(eq_attr "alternative" "17,18")
(const_string "ssecvt")
(eq_attr "alternative" "4")
(const_string "multi")
(match_operand:DI 1 "pic_32bit_operand" "")
(const_string "lea")
]
(const_string "imov")))
(set (attr "modrm")
(if_then_else
(and (eq_attr "alternative" "2") (eq_attr "type" "imov"))
(const_string "0")
(const_string "*")))
(set (attr "length_immediate")
(if_then_else
(and (eq_attr "alternative" "2") (eq_attr "type" "imov"))
(const_string "8")
(const_string "*")))
(set_attr "prefix_rex" "*,*,*,*,*,*,*,1,*,1,*,*,*,*,*,*,*,*,*")
(set_attr "prefix_data16" "*,*,*,*,*,*,*,*,*,*,*,*,*,*,*,1,*,*,*")
(set (attr "prefix")
(if_then_else (eq_attr "alternative" "11,12,13,14,15,16")
(const_string "maybe_vex")
(const_string "orig")))
(set_attr "mode" "SI,DI,DI,DI,SI,DI,DI,DI,DI,DI,DI,TI,TI,DI,DI,DI,DI,DI,DI")])
;; Stores and loads of ax to arbitrary constant address.
;; We fake an second form of instruction to force reload to load address
;; into register when rax is not available
(define_insn "*movabsdi_1_rex64"
[(set (mem:DI (match_operand:DI 0 "x86_64_movabs_operand" "i,r"))
(match_operand:DI 1 "nonmemory_operand" "a,er"))]
"TARGET_64BIT && ix86_check_movabs (insn, 0)"
"@
movabs{q}\t{%1, %P0|%P0, %1}
mov{q}\t{%1, %a0|%a0, %1}"
[(set_attr "type" "imov")
(set_attr "modrm" "0,*")
(set_attr "length_address" "8,0")
(set_attr "length_immediate" "0,*")
(set_attr "memory" "store")
(set_attr "mode" "DI")])
(define_insn "*movabsdi_2_rex64"
[(set (match_operand:DI 0 "register_operand" "=a,r")
(mem:DI (match_operand:DI 1 "x86_64_movabs_operand" "i,r")))]
"TARGET_64BIT && ix86_check_movabs (insn, 1)"
"@
movabs{q}\t{%P1, %0|%0, %P1}
mov{q}\t{%a1, %0|%0, %a1}"
[(set_attr "type" "imov")
(set_attr "modrm" "0,*")
(set_attr "length_address" "8,0")
(set_attr "length_immediate" "0")
(set_attr "memory" "load")
(set_attr "mode" "DI")])
;; Convert impossible stores of immediate to existing instructions.
;; First try to get scratch register and go through it. In case this
;; fails, move by 32bit parts.
(define_peephole2
[(match_scratch:DI 2 "r")
(set (match_operand:DI 0 "memory_operand" "")
(match_operand:DI 1 "immediate_operand" ""))]
"TARGET_64BIT && !symbolic_operand (operands[1], DImode)
&& !x86_64_immediate_operand (operands[1], DImode)"
[(set (match_dup 2) (match_dup 1))
(set (match_dup 0) (match_dup 2))]
"")
;; We need to define this as both peepholer and splitter for case
;; peephole2 pass is not run.
;; "&& 1" is needed to keep it from matching the previous pattern.
(define_peephole2
[(set (match_operand:DI 0 "memory_operand" "")
(match_operand:DI 1 "immediate_operand" ""))]
"TARGET_64BIT && !symbolic_operand (operands[1], DImode)
&& !x86_64_immediate_operand (operands[1], DImode) && 1"
[(set (match_dup 2) (match_dup 3))
(set (match_dup 4) (match_dup 5))]
"split_di (&operands[0], 2, &operands[2], &operands[4]);")
(define_split
[(set (match_operand:DI 0 "memory_operand" "")
(match_operand:DI 1 "immediate_operand" ""))]
"TARGET_64BIT && ((optimize > 0 && flag_peephole2)
? epilogue_completed : reload_completed)
&& !symbolic_operand (operands[1], DImode)
&& !x86_64_immediate_operand (operands[1], DImode)"
[(set (match_dup 2) (match_dup 3))
(set (match_dup 4) (match_dup 5))]
"split_di (&operands[0], 2, &operands[2], &operands[4]);")
(define_insn "*swapdi_rex64"
[(set (match_operand:DI 0 "register_operand" "+r")
(match_operand:DI 1 "register_operand" "+r"))
(set (match_dup 1)
(match_dup 0))]
"TARGET_64BIT"
"xchg{q}\t%1, %0"
[(set_attr "type" "imov")
(set_attr "mode" "DI")
(set_attr "pent_pair" "np")
(set_attr "athlon_decode" "vector")
(set_attr "amdfam10_decode" "double")])
(define_expand "movoi"
[(set (match_operand:OI 0 "nonimmediate_operand" "")
(match_operand:OI 1 "general_operand" ""))]
"TARGET_AVX"
"ix86_expand_move (OImode, operands); DONE;")
(define_insn "*movoi_internal"
[(set (match_operand:OI 0 "nonimmediate_operand" "=x,x,m")
(match_operand:OI 1 "vector_move_operand" "C,xm,x"))]
"TARGET_AVX
&& !(MEM_P (operands[0]) && MEM_P (operands[1]))"
{
switch (which_alternative)
{
case 0:
return "vxorps\t%0, %0, %0";
case 1:
case 2:
if (misaligned_operand (operands[0], OImode)
|| misaligned_operand (operands[1], OImode))
return "vmovdqu\t{%1, %0|%0, %1}";
else
return "vmovdqa\t{%1, %0|%0, %1}";
default:
gcc_unreachable ();
}
}
[(set_attr "type" "sselog1,ssemov,ssemov")
(set_attr "prefix" "vex")
(set_attr "mode" "OI")])
(define_expand "movti"
[(set (match_operand:TI 0 "nonimmediate_operand" "")
(match_operand:TI 1 "nonimmediate_operand" ""))]
"TARGET_SSE || TARGET_64BIT"
{
if (TARGET_64BIT)
ix86_expand_move (TImode, operands);
else if (push_operand (operands[0], TImode))
ix86_expand_push (TImode, operands[1]);
else
ix86_expand_vector_move (TImode, operands);
DONE;
})
(define_insn "*movti_internal"
[(set (match_operand:TI 0 "nonimmediate_operand" "=x,x,m")
(match_operand:TI 1 "vector_move_operand" "C,xm,x"))]
"TARGET_SSE && !TARGET_64BIT
&& !(MEM_P (operands[0]) && MEM_P (operands[1]))"
{
switch (which_alternative)
{
case 0:
if (get_attr_mode (insn) == MODE_V4SF)
return "%vxorps\t%0, %d0";
else
return "%vpxor\t%0, %d0";
case 1:
case 2:
/* TDmode values are passed as TImode on the stack. Moving them
to stack may result in unaligned memory access. */
if (misaligned_operand (operands[0], TImode)
|| misaligned_operand (operands[1], TImode))
{
if (get_attr_mode (insn) == MODE_V4SF)
return "%vmovups\t{%1, %0|%0, %1}";
else
return "%vmovdqu\t{%1, %0|%0, %1}";
}
else
{
if (get_attr_mode (insn) == MODE_V4SF)
return "%vmovaps\t{%1, %0|%0, %1}";
else
return "%vmovdqa\t{%1, %0|%0, %1}";
}
default:
gcc_unreachable ();
}
}
[(set_attr "type" "sselog1,ssemov,ssemov")
(set_attr "prefix" "maybe_vex")
(set (attr "mode")
(cond [(ior (eq (symbol_ref "TARGET_SSE2") (const_int 0))
(ne (symbol_ref "optimize_function_for_size_p (cfun)") (const_int 0)))
(const_string "V4SF")
(and (eq_attr "alternative" "2")
(ne (symbol_ref "TARGET_SSE_TYPELESS_STORES")
(const_int 0)))
(const_string "V4SF")]
(const_string "TI")))])
(define_insn "*movti_rex64"
[(set (match_operand:TI 0 "nonimmediate_operand" "=!r,o,x,x,xm")
(match_operand:TI 1 "general_operand" "riFo,riF,C,xm,x"))]
"TARGET_64BIT
&& !(MEM_P (operands[0]) && MEM_P (operands[1]))"
{
switch (which_alternative)
{
case 0:
case 1:
return "#";
case 2:
if (get_attr_mode (insn) == MODE_V4SF)
return "%vxorps\t%0, %d0";
else
return "%vpxor\t%0, %d0";
case 3:
case 4:
/* TDmode values are passed as TImode on the stack. Moving them
to stack may result in unaligned memory access. */
if (misaligned_operand (operands[0], TImode)
|| misaligned_operand (operands[1], TImode))
{
if (get_attr_mode (insn) == MODE_V4SF)
return "%vmovups\t{%1, %0|%0, %1}";
else
return "%vmovdqu\t{%1, %0|%0, %1}";
}
else
{
if (get_attr_mode (insn) == MODE_V4SF)
return "%vmovaps\t{%1, %0|%0, %1}";
else
return "%vmovdqa\t{%1, %0|%0, %1}";
}
default:
gcc_unreachable ();
}
}
[(set_attr "type" "*,*,sselog1,ssemov,ssemov")
(set_attr "prefix" "*,*,maybe_vex,maybe_vex,maybe_vex")
(set (attr "mode")
(cond [(eq_attr "alternative" "2,3")
(if_then_else
(ne (symbol_ref "optimize_function_for_size_p (cfun)")
(const_int 0))
(const_string "V4SF")
(const_string "TI"))
(eq_attr "alternative" "4")
(if_then_else
(ior (ne (symbol_ref "TARGET_SSE_TYPELESS_STORES")
(const_int 0))
(ne (symbol_ref "optimize_function_for_size_p (cfun)")
(const_int 0)))
(const_string "V4SF")
(const_string "TI"))]
(const_string "DI")))])
(define_split
[(set (match_operand:TI 0 "nonimmediate_operand" "")
(match_operand:TI 1 "general_operand" ""))]
"reload_completed && !SSE_REG_P (operands[0])
&& !SSE_REG_P (operands[1])"
[(const_int 0)]
"ix86_split_long_move (operands); DONE;")
;; This expands to what emit_move_complex would generate if we didn't
;; have a movti pattern. Having this avoids problems with reload on
;; 32-bit targets when SSE is present, but doesn't seem to be harmful
;; to have around all the time.
(define_expand "movcdi"
[(set (match_operand:CDI 0 "nonimmediate_operand" "")
(match_operand:CDI 1 "general_operand" ""))]
""
{
if (push_operand (operands[0], CDImode))
emit_move_complex_push (CDImode, operands[0], operands[1]);
else
emit_move_complex_parts (operands[0], operands[1]);
DONE;
})
(define_expand "movsf"
[(set (match_operand:SF 0 "nonimmediate_operand" "")
(match_operand:SF 1 "general_operand" ""))]
""
"ix86_expand_move (SFmode, operands); DONE;")
(define_insn "*pushsf"
[(set (match_operand:SF 0 "push_operand" "=<,<,<")
(match_operand:SF 1 "general_no_elim_operand" "f,rFm,x"))]
"!TARGET_64BIT"
{
/* Anything else should be already split before reg-stack. */
gcc_assert (which_alternative == 1);
return "push{l}\t%1";
}
[(set_attr "type" "multi,push,multi")
(set_attr "unit" "i387,*,*")
(set_attr "mode" "SF,SI,SF")])
(define_insn "*pushsf_rex64"
[(set (match_operand:SF 0 "push_operand" "=X,X,X")
(match_operand:SF 1 "nonmemory_no_elim_operand" "f,rF,x"))]
"TARGET_64BIT"
{
/* Anything else should be already split before reg-stack. */
gcc_assert (which_alternative == 1);
return "push{q}\t%q1";
}
[(set_attr "type" "multi,push,multi")
(set_attr "unit" "i387,*,*")
(set_attr "mode" "SF,DI,SF")])
(define_split
[(set (match_operand:SF 0 "push_operand" "")
(match_operand:SF 1 "memory_operand" ""))]
"reload_completed
&& MEM_P (operands[1])
&& (operands[2] = find_constant_src (insn))"
[(set (match_dup 0)
(match_dup 2))])
;; %%% Kill this when call knows how to work this out.
(define_split
[(set (match_operand:SF 0 "push_operand" "")
(match_operand:SF 1 "any_fp_register_operand" ""))]
"!TARGET_64BIT"
[(set (reg:SI SP_REG) (plus:SI (reg:SI SP_REG) (const_int -4)))
(set (mem:SF (reg:SI SP_REG)) (match_dup 1))])
(define_split
[(set (match_operand:SF 0 "push_operand" "")
(match_operand:SF 1 "any_fp_register_operand" ""))]
"TARGET_64BIT"
[(set (reg:DI SP_REG) (plus:DI (reg:DI SP_REG) (const_int -8)))
(set (mem:SF (reg:DI SP_REG)) (match_dup 1))])
(define_insn "*movsf_1"
[(set (match_operand:SF 0 "nonimmediate_operand"
"=f,m,f,r ,m ,x,x,x ,m,!*y,!m,!*y,?Yi,?r,!*Ym,!r")
(match_operand:SF 1 "general_operand"
"fm,f,G,rmF,Fr,C,x,xm,x,m ,*y,*y ,r ,Yi,r ,*Ym"))]
"!(MEM_P (operands[0]) && MEM_P (operands[1]))
&& (reload_in_progress || reload_completed
|| (ix86_cmodel == CM_MEDIUM || ix86_cmodel == CM_LARGE)
|| (!TARGET_SSE_MATH && optimize_function_for_size_p (cfun)
&& standard_80387_constant_p (operands[1]))
|| GET_CODE (operands[1]) != CONST_DOUBLE
|| memory_operand (operands[0], SFmode))"
{
switch (which_alternative)
{
case 0:
case 1:
return output_387_reg_move (insn, operands);
case 2:
return standard_80387_constant_opcode (operands[1]);
case 3:
case 4:
return "mov{l}\t{%1, %0|%0, %1}";
case 5:
if (get_attr_mode (insn) == MODE_TI)
return "%vpxor\t%0, %d0";
else
return "%vxorps\t%0, %d0";
case 6:
if (get_attr_mode (insn) == MODE_V4SF)
return "%vmovaps\t{%1, %0|%0, %1}";
else
return "%vmovss\t{%1, %d0|%d0, %1}";
case 7:
if (TARGET_AVX)
return REG_P (operands[1]) ? "vmovss\t{%1, %0, %0|%0, %0, %1}"
: "vmovss\t{%1, %0|%0, %1}";
else
return "movss\t{%1, %0|%0, %1}";
case 8:
return "%vmovss\t{%1, %0|%0, %1}";
case 9: case 10: case 14: case 15:
return "movd\t{%1, %0|%0, %1}";
case 12: case 13:
return "%vmovd\t{%1, %0|%0, %1}";
case 11:
return "movq\t{%1, %0|%0, %1}";
default:
gcc_unreachable ();
}
}
[(set_attr "type" "fmov,fmov,fmov,imov,imov,sselog1,ssemov,ssemov,ssemov,mmxmov,mmxmov,mmxmov,ssemov,ssemov,mmxmov,mmxmov")
(set (attr "prefix")
(if_then_else (eq_attr "alternative" "5,6,7,8,12,13")
(const_string "maybe_vex")
(const_string "orig")))
(set (attr "mode")
(cond [(eq_attr "alternative" "3,4,9,10")
(const_string "SI")
(eq_attr "alternative" "5")
(if_then_else
(and (and (ne (symbol_ref "TARGET_SSE_LOAD0_BY_PXOR")
(const_int 0))
(ne (symbol_ref "TARGET_SSE2")
(const_int 0)))
(eq (symbol_ref "optimize_function_for_size_p (cfun)")
(const_int 0)))
(const_string "TI")
(const_string "V4SF"))
/* For architectures resolving dependencies on
whole SSE registers use APS move to break dependency
chains, otherwise use short move to avoid extra work.
Do the same for architectures resolving dependencies on
the parts. While in DF mode it is better to always handle
just register parts, the SF mode is different due to lack
of instructions to load just part of the register. It is
better to maintain the whole registers in single format
to avoid problems on using packed logical operations. */
(eq_attr "alternative" "6")
(if_then_else
(ior (ne (symbol_ref "TARGET_SSE_PARTIAL_REG_DEPENDENCY")
(const_int 0))
(ne (symbol_ref "TARGET_SSE_SPLIT_REGS")
(const_int 0)))
(const_string "V4SF")
(const_string "SF"))
(eq_attr "alternative" "11")
(const_string "DI")]
(const_string "SF")))])
(define_insn "*swapsf"
[(set (match_operand:SF 0 "fp_register_operand" "+f")
(match_operand:SF 1 "fp_register_operand" "+f"))
(set (match_dup 1)
(match_dup 0))]
"reload_completed || TARGET_80387"
{
if (STACK_TOP_P (operands[0]))
return "fxch\t%1";
else
return "fxch\t%0";
}
[(set_attr "type" "fxch")
(set_attr "mode" "SF")])
(define_expand "movdf"
[(set (match_operand:DF 0 "nonimmediate_operand" "")
(match_operand:DF 1 "general_operand" ""))]
""
"ix86_expand_move (DFmode, operands); DONE;")
;; Size of pushdf is 3 (for sub) + 2 (for fstp) + memory operand size.
;; Size of pushdf using integer instructions is 2+2*memory operand size
;; On the average, pushdf using integers can be still shorter. Allow this
;; pattern for optimize_size too.
(define_insn "*pushdf_nointeger"
[(set (match_operand:DF 0 "push_operand" "=<,<,<,<")
(match_operand:DF 1 "general_no_elim_operand" "f,Fo,*r,Y2"))]
"!TARGET_64BIT && !TARGET_INTEGER_DFMODE_MOVES"
{
/* This insn should be already split before reg-stack. */
gcc_unreachable ();
}
[(set_attr "type" "multi")
(set_attr "unit" "i387,*,*,*")
(set_attr "mode" "DF,SI,SI,DF")])
(define_insn "*pushdf_integer"
[(set (match_operand:DF 0 "push_operand" "=<,<,<")
(match_operand:DF 1 "general_no_elim_operand" "f,rFo,Y2"))]
"TARGET_64BIT || TARGET_INTEGER_DFMODE_MOVES"
{
/* This insn should be already split before reg-stack. */
gcc_unreachable ();
}
[(set_attr "type" "multi")
(set_attr "unit" "i387,*,*")
(set_attr "mode" "DF,SI,DF")])
;; %%% Kill this when call knows how to work this out.
(define_split
[(set (match_operand:DF 0 "push_operand" "")
(match_operand:DF 1 "any_fp_register_operand" ""))]
"reload_completed"
[(set (reg:P SP_REG) (plus:P (reg:P SP_REG) (const_int -8)))
(set (mem:DF (reg:P SP_REG)) (match_dup 1))]
"")
(define_split
[(set (match_operand:DF 0 "push_operand" "")
(match_operand:DF 1 "general_operand" ""))]
"reload_completed"
[(const_int 0)]
"ix86_split_long_move (operands); DONE;")
;; Moving is usually shorter when only FP registers are used. This separate
;; movdf pattern avoids the use of integer registers for FP operations
;; when optimizing for size.
(define_insn "*movdf_nointeger"
[(set (match_operand:DF 0 "nonimmediate_operand"
"=f,m,f,*r ,o ,Y2*x,Y2*x,Y2*x ,m ")
(match_operand:DF 1 "general_operand"
"fm,f,G,*roF,*Fr,C ,Y2*x,mY2*x,Y2*x"))]
"!(MEM_P (operands[0]) && MEM_P (operands[1]))
&& ((optimize_function_for_size_p (cfun)
|| !TARGET_INTEGER_DFMODE_MOVES) && !TARGET_64BIT)
&& (reload_in_progress || reload_completed
|| (ix86_cmodel == CM_MEDIUM || ix86_cmodel == CM_LARGE)
|| (!(TARGET_SSE2 && TARGET_SSE_MATH)
&& optimize_function_for_size_p (cfun)
&& !memory_operand (operands[0], DFmode)
&& standard_80387_constant_p (operands[1]))
|| GET_CODE (operands[1]) != CONST_DOUBLE
|| ((optimize_function_for_size_p (cfun)
|| !TARGET_MEMORY_MISMATCH_STALL
|| reload_in_progress || reload_completed)
&& memory_operand (operands[0], DFmode)))"
{
switch (which_alternative)
{
case 0:
case 1:
return output_387_reg_move (insn, operands);
case 2:
return standard_80387_constant_opcode (operands[1]);
case 3:
case 4:
return "#";
case 5:
switch (get_attr_mode (insn))
{
case MODE_V4SF:
return "%vxorps\t%0, %d0";
case MODE_V2DF:
return "%vxorpd\t%0, %d0";
case MODE_TI:
return "%vpxor\t%0, %d0";
default:
gcc_unreachable ();
}
case 6:
case 7:
case 8:
switch (get_attr_mode (insn))
{
case MODE_V4SF:
return "%vmovaps\t{%1, %0|%0, %1}";
case MODE_V2DF:
return "%vmovapd\t{%1, %0|%0, %1}";
case MODE_TI:
return "%vmovdqa\t{%1, %0|%0, %1}";
case MODE_DI:
return "%vmovq\t{%1, %0|%0, %1}";
case MODE_DF:
if (TARGET_AVX)
{
if (REG_P (operands[0]) && REG_P (operands[1]))
return "vmovsd\t{%1, %0, %0|%0, %0, %1}";
else
return "vmovsd\t{%1, %0|%0, %1}";
}
else
return "movsd\t{%1, %0|%0, %1}";
case MODE_V1DF:
if (TARGET_AVX)
{
if (REG_P (operands[0]))
return "vmovlpd\t{%1, %0, %0|%0, %0, %1}";
else
return "vmovlpd\t{%1, %0|%0, %1}";
}
else
return "movlpd\t{%1, %0|%0, %1}";
case MODE_V2SF:
if (TARGET_AVX)
{
if (REG_P (operands[0]))
return "vmovlps\t{%1, %0, %0|%0, %0, %1}";
else
return "vmovlps\t{%1, %0|%0, %1}";
}
else
return "movlps\t{%1, %0|%0, %1}";
default:
gcc_unreachable ();
}
default:
gcc_unreachable ();
}
}
[(set_attr "type" "fmov,fmov,fmov,multi,multi,sselog1,ssemov,ssemov,ssemov")
(set (attr "prefix")
(if_then_else (eq_attr "alternative" "0,1,2,3,4")
(const_string "orig")
(const_string "maybe_vex")))
(set (attr "prefix_data16")
(if_then_else (eq_attr "mode" "V1DF")
(const_string "1")
(const_string "*")))
(set (attr "mode")
(cond [(eq_attr "alternative" "0,1,2")
(const_string "DF")
(eq_attr "alternative" "3,4")
(const_string "SI")
/* For SSE1, we have many fewer alternatives. */
(eq (symbol_ref "TARGET_SSE2") (const_int 0))
(cond [(eq_attr "alternative" "5,6")
(const_string "V4SF")
]
(const_string "V2SF"))
/* xorps is one byte shorter. */
(eq_attr "alternative" "5")
(cond [(ne (symbol_ref "optimize_function_for_size_p (cfun)")
(const_int 0))
(const_string "V4SF")
(ne (symbol_ref "TARGET_SSE_LOAD0_BY_PXOR")
(const_int 0))
(const_string "TI")
]
(const_string "V2DF"))
/* For architectures resolving dependencies on
whole SSE registers use APD move to break dependency
chains, otherwise use short move to avoid extra work.
movaps encodes one byte shorter. */
(eq_attr "alternative" "6")
(cond
[(ne (symbol_ref "optimize_function_for_size_p (cfun)")
(const_int 0))
(const_string "V4SF")
(ne (symbol_ref "TARGET_SSE_PARTIAL_REG_DEPENDENCY")
(const_int 0))
(const_string "V2DF")
]
(const_string "DF"))
/* For architectures resolving dependencies on register
parts we may avoid extra work to zero out upper part
of register. */
(eq_attr "alternative" "7")
(if_then_else
(ne (symbol_ref "TARGET_SSE_SPLIT_REGS")
(const_int 0))
(const_string "V1DF")
(const_string "DF"))
]
(const_string "DF")))])
(define_insn "*movdf_integer_rex64"
[(set (match_operand:DF 0 "nonimmediate_operand"
"=f,m,f,r ,m ,Y2*x,Y2*x,Y2*x,m ,Yi,r ")
(match_operand:DF 1 "general_operand"
"fm,f,G,rmF,Fr,C ,Y2*x,m ,Y2*x,r ,Yi"))]
"TARGET_64BIT && !(MEM_P (operands[0]) && MEM_P (operands[1]))
&& (reload_in_progress || reload_completed
|| (ix86_cmodel == CM_MEDIUM || ix86_cmodel == CM_LARGE)
|| (!(TARGET_SSE2 && TARGET_SSE_MATH)
&& optimize_function_for_size_p (cfun)
&& standard_80387_constant_p (operands[1]))
|| GET_CODE (operands[1]) != CONST_DOUBLE
|| memory_operand (operands[0], DFmode))"
{
switch (which_alternative)
{
case 0:
case 1:
return output_387_reg_move (insn, operands);
case 2:
return standard_80387_constant_opcode (operands[1]);
case 3:
case 4:
return "#";
case 5:
switch (get_attr_mode (insn))
{
case MODE_V4SF:
return "%vxorps\t%0, %d0";
case MODE_V2DF:
return "%vxorpd\t%0, %d0";
case MODE_TI:
return "%vpxor\t%0, %d0";
default:
gcc_unreachable ();
}
case 6:
case 7:
case 8:
switch (get_attr_mode (insn))
{
case MODE_V4SF:
return "%vmovaps\t{%1, %0|%0, %1}";
case MODE_V2DF:
return "%vmovapd\t{%1, %0|%0, %1}";
case MODE_TI:
return "%vmovdqa\t{%1, %0|%0, %1}";
case MODE_DI:
return "%vmovq\t{%1, %0|%0, %1}";
case MODE_DF:
if (TARGET_AVX)
{
if (REG_P (operands[0]) && REG_P (operands[1]))
return "vmovsd\t{%1, %0, %0|%0, %0, %1}";
else
return "vmovsd\t{%1, %0|%0, %1}";
}
else
return "movsd\t{%1, %0|%0, %1}";
case MODE_V1DF:
return "%vmovlpd\t{%1, %d0|%d0, %1}";
case MODE_V2SF:
return "%vmovlps\t{%1, %d0|%d0, %1}";
default:
gcc_unreachable ();
}
case 9:
case 10:
return "%vmovd\t{%1, %0|%0, %1}";
default:
gcc_unreachable();
}
}
[(set_attr "type" "fmov,fmov,fmov,multi,multi,sselog1,ssemov,ssemov,ssemov,ssemov,ssemov")
(set (attr "prefix")
(if_then_else (eq_attr "alternative" "0,1,2,3,4")
(const_string "orig")
(const_string "maybe_vex")))
(set (attr "prefix_data16")
(if_then_else (eq_attr "mode" "V1DF")
(const_string "1")
(const_string "*")))
(set (attr "mode")
(cond [(eq_attr "alternative" "0,1,2")
(const_string "DF")
(eq_attr "alternative" "3,4,9,10")
(const_string "DI")
/* For SSE1, we have many fewer alternatives. */
(eq (symbol_ref "TARGET_SSE2") (const_int 0))
(cond [(eq_attr "alternative" "5,6")
(const_string "V4SF")
]
(const_string "V2SF"))
/* xorps is one byte shorter. */
(eq_attr "alternative" "5")
(cond [(ne (symbol_ref "optimize_function_for_size_p (cfun)")
(const_int 0))
(const_string "V4SF")
(ne (symbol_ref "TARGET_SSE_LOAD0_BY_PXOR")
(const_int 0))
(const_string "TI")
]
(const_string "V2DF"))
/* For architectures resolving dependencies on
whole SSE registers use APD move to break dependency
chains, otherwise use short move to avoid extra work.
movaps encodes one byte shorter. */
(eq_attr "alternative" "6")
(cond
[(ne (symbol_ref "optimize_function_for_size_p (cfun)")
(const_int 0))
(const_string "V4SF")
(ne (symbol_ref "TARGET_SSE_PARTIAL_REG_DEPENDENCY")
(const_int 0))
(const_string "V2DF")
]
(const_string "DF"))
/* For architectures resolving dependencies on register
parts we may avoid extra work to zero out upper part
of register. */
(eq_attr "alternative" "7")
(if_then_else
(ne (symbol_ref "TARGET_SSE_SPLIT_REGS")
(const_int 0))
(const_string "V1DF")
(const_string "DF"))
]
(const_string "DF")))])
(define_insn "*movdf_integer"
[(set (match_operand:DF 0 "nonimmediate_operand"
"=f,m,f,r ,o ,Y2*x,Y2*x,Y2*x,m ")
(match_operand:DF 1 "general_operand"
"fm,f,G,roF,Fr,C ,Y2*x,m ,Y2*x"))]
"!(MEM_P (operands[0]) && MEM_P (operands[1]))
&& optimize_function_for_speed_p (cfun)
&& TARGET_INTEGER_DFMODE_MOVES
&& (reload_in_progress || reload_completed
|| (ix86_cmodel == CM_MEDIUM || ix86_cmodel == CM_LARGE)
|| (!(TARGET_SSE2 && TARGET_SSE_MATH)
&& optimize_function_for_size_p (cfun)
&& standard_80387_constant_p (operands[1]))
|| GET_CODE (operands[1]) != CONST_DOUBLE
|| memory_operand (operands[0], DFmode))"
{
switch (which_alternative)
{
case 0:
case 1:
return output_387_reg_move (insn, operands);
case 2:
return standard_80387_constant_opcode (operands[1]);
case 3:
case 4:
return "#";
case 5:
switch (get_attr_mode (insn))
{
case MODE_V4SF:
return "xorps\t%0, %0";
case MODE_V2DF:
return "xorpd\t%0, %0";
case MODE_TI:
return "pxor\t%0, %0";
default:
gcc_unreachable ();
}
case 6:
case 7:
case 8:
switch (get_attr_mode (insn))
{
case MODE_V4SF:
return "movaps\t{%1, %0|%0, %1}";
case MODE_V2DF:
return "movapd\t{%1, %0|%0, %1}";
case MODE_TI:
return "movdqa\t{%1, %0|%0, %1}";
case MODE_DI:
return "movq\t{%1, %0|%0, %1}";
case MODE_DF:
return "movsd\t{%1, %0|%0, %1}";
case MODE_V1DF:
return "movlpd\t{%1, %0|%0, %1}";
case MODE_V2SF:
return "movlps\t{%1, %0|%0, %1}";
default:
gcc_unreachable ();
}
default:
gcc_unreachable();
}
}
[(set_attr "type" "fmov,fmov,fmov,multi,multi,sselog1,ssemov,ssemov,ssemov")
(set (attr "prefix_data16")
(if_then_else (eq_attr "mode" "V1DF")
(const_string "1")
(const_string "*")))
(set (attr "mode")
(cond [(eq_attr "alternative" "0,1,2")
(const_string "DF")
(eq_attr "alternative" "3,4")
(const_string "SI")
/* For SSE1, we have many fewer alternatives. */
(eq (symbol_ref "TARGET_SSE2") (const_int 0))
(cond [(eq_attr "alternative" "5,6")
(const_string "V4SF")
]
(const_string "V2SF"))
/* xorps is one byte shorter. */
(eq_attr "alternative" "5")
(cond [(ne (symbol_ref "optimize_function_for_size_p (cfun)")
(const_int 0))
(const_string "V4SF")
(ne (symbol_ref "TARGET_SSE_LOAD0_BY_PXOR")
(const_int 0))
(const_string "TI")
]
(const_string "V2DF"))
/* For architectures resolving dependencies on
whole SSE registers use APD move to break dependency
chains, otherwise use short move to avoid extra work.
movaps encodes one byte shorter. */
(eq_attr "alternative" "6")
(cond
[(ne (symbol_ref "optimize_function_for_size_p (cfun)")
(const_int 0))
(const_string "V4SF")
(ne (symbol_ref "TARGET_SSE_PARTIAL_REG_DEPENDENCY")
(const_int 0))
(const_string "V2DF")
]
(const_string "DF"))
/* For architectures resolving dependencies on register
parts we may avoid extra work to zero out upper part
of register. */
(eq_attr "alternative" "7")
(if_then_else
(ne (symbol_ref "TARGET_SSE_SPLIT_REGS")
(const_int 0))
(const_string "V1DF")
(const_string "DF"))
]
(const_string "DF")))])
(define_split
[(set (match_operand:DF 0 "nonimmediate_operand" "")
(match_operand:DF 1 "general_operand" ""))]
"reload_completed
&& !(MEM_P (operands[0]) && MEM_P (operands[1]))
&& ! (ANY_FP_REG_P (operands[0]) ||
(GET_CODE (operands[0]) == SUBREG
&& ANY_FP_REG_P (SUBREG_REG (operands[0]))))
&& ! (ANY_FP_REG_P (operands[1]) ||
(GET_CODE (operands[1]) == SUBREG
&& ANY_FP_REG_P (SUBREG_REG (operands[1]))))"
[(const_int 0)]
"ix86_split_long_move (operands); DONE;")
(define_insn "*swapdf"
[(set (match_operand:DF 0 "fp_register_operand" "+f")
(match_operand:DF 1 "fp_register_operand" "+f"))
(set (match_dup 1)
(match_dup 0))]
"reload_completed || TARGET_80387"
{
if (STACK_TOP_P (operands[0]))
return "fxch\t%1";
else
return "fxch\t%0";
}
[(set_attr "type" "fxch")
(set_attr "mode" "DF")])
(define_expand "movxf"
[(set (match_operand:XF 0 "nonimmediate_operand" "")
(match_operand:XF 1 "general_operand" ""))]
""
"ix86_expand_move (XFmode, operands); DONE;")
;; Size of pushdf is 3 (for sub) + 2 (for fstp) + memory operand size.
;; Size of pushdf using integer instructions is 3+3*memory operand size
;; Pushing using integer instructions is longer except for constants
;; and direct memory references.
;; (assuming that any given constant is pushed only once, but this ought to be
;; handled elsewhere).
(define_insn "*pushxf_nointeger"
[(set (match_operand:XF 0 "push_operand" "=X,X,X")
(match_operand:XF 1 "general_no_elim_operand" "f,Fo,*r"))]
"optimize_function_for_size_p (cfun)"
{
/* This insn should be already split before reg-stack. */
gcc_unreachable ();
}
[(set_attr "type" "multi")
(set_attr "unit" "i387,*,*")
(set_attr "mode" "XF,SI,SI")])
(define_insn "*pushxf_integer"
[(set (match_operand:XF 0 "push_operand" "=<,<")
(match_operand:XF 1 "general_no_elim_operand" "f,ro"))]
"optimize_function_for_speed_p (cfun)"
{
/* This insn should be already split before reg-stack. */
gcc_unreachable ();
}
[(set_attr "type" "multi")
(set_attr "unit" "i387,*")
(set_attr "mode" "XF,SI")])
(define_split
[(set (match_operand 0 "push_operand" "")
(match_operand 1 "general_operand" ""))]
"reload_completed
&& (GET_MODE (operands[0]) == XFmode
|| GET_MODE (operands[0]) == DFmode)
&& !ANY_FP_REG_P (operands[1])"
[(const_int 0)]
"ix86_split_long_move (operands); DONE;")
(define_split
[(set (match_operand:XF 0 "push_operand" "")
(match_operand:XF 1 "any_fp_register_operand" ""))]
""
[(set (reg:P SP_REG) (plus:P (reg:P SP_REG) (match_dup 2)))
(set (mem:XF (reg:P SP_REG)) (match_dup 1))]
"operands[2] = GEN_INT (TARGET_128BIT_LONG_DOUBLE ? -16 : -12);")
;; Do not use integer registers when optimizing for size
(define_insn "*movxf_nointeger"
[(set (match_operand:XF 0 "nonimmediate_operand" "=f,m,f,*r,o")
(match_operand:XF 1 "general_operand" "fm,f,G,*roF,F*r"))]
"optimize_function_for_size_p (cfun)
&& !(MEM_P (operands[0]) && MEM_P (operands[1]))
&& (reload_in_progress || reload_completed
|| standard_80387_constant_p (operands[1])
|| GET_CODE (operands[1]) != CONST_DOUBLE
|| memory_operand (operands[0], XFmode))"
{
switch (which_alternative)
{
case 0:
case 1:
return output_387_reg_move (insn, operands);
case 2:
return standard_80387_constant_opcode (operands[1]);
case 3: case 4:
return "#";
default:
gcc_unreachable ();
}
}
[(set_attr "type" "fmov,fmov,fmov,multi,multi")
(set_attr "mode" "XF,XF,XF,SI,SI")])
(define_insn "*movxf_integer"
[(set (match_operand:XF 0 "nonimmediate_operand" "=f,m,f,r,o")
(match_operand:XF 1 "general_operand" "fm,f,G,roF,Fr"))]
"optimize_function_for_speed_p (cfun)
&& !(MEM_P (operands[0]) && MEM_P (operands[1]))
&& (reload_in_progress || reload_completed
|| GET_CODE (operands[1]) != CONST_DOUBLE
|| memory_operand (operands[0], XFmode))"
{
switch (which_alternative)
{
case 0:
case 1:
return output_387_reg_move (insn, operands);
case 2:
return standard_80387_constant_opcode (operands[1]);
case 3: case 4:
return "#";
default:
gcc_unreachable ();
}
}
[(set_attr "type" "fmov,fmov,fmov,multi,multi")
(set_attr "mode" "XF,XF,XF,SI,SI")])
(define_expand "movtf"
[(set (match_operand:TF 0 "nonimmediate_operand" "")
(match_operand:TF 1 "nonimmediate_operand" ""))]
"TARGET_SSE2"
{
ix86_expand_move (TFmode, operands);
DONE;
})
(define_insn "*movtf_internal"
[(set (match_operand:TF 0 "nonimmediate_operand" "=x,m,x,?r,?o")
(match_operand:TF 1 "general_operand" "xm,x,C,roF,Fr"))]
"TARGET_SSE2
&& !(MEM_P (operands[0]) && MEM_P (operands[1]))"
{
switch (which_alternative)
{
case 0:
case 1:
if (get_attr_mode (insn) == MODE_V4SF)
return "%vmovaps\t{%1, %0|%0, %1}";
else
return "%vmovdqa\t{%1, %0|%0, %1}";
case 2:
if (get_attr_mode (insn) == MODE_V4SF)
return "%vxorps\t%0, %d0";
else
return "%vpxor\t%0, %d0";
case 3:
case 4:
return "#";
default:
gcc_unreachable ();
}
}
[(set_attr "type" "ssemov,ssemov,sselog1,*,*")
(set_attr "prefix" "maybe_vex,maybe_vex,maybe_vex,*,*")
(set (attr "mode")
(cond [(eq_attr "alternative" "0,2")
(if_then_else
(ne (symbol_ref "optimize_function_for_size_p (cfun)")
(const_int 0))
(const_string "V4SF")
(const_string "TI"))
(eq_attr "alternative" "1")
(if_then_else
(ior (ne (symbol_ref "TARGET_SSE_TYPELESS_STORES")
(const_int 0))
(ne (symbol_ref "optimize_function_for_size_p (cfun)")
(const_int 0)))
(const_string "V4SF")
(const_string "TI"))]
(const_string "DI")))])
(define_insn "*pushtf_sse"
[(set (match_operand:TF 0 "push_operand" "=<,<,<")
(match_operand:TF 1 "general_no_elim_operand" "x,Fo,*r"))]
"TARGET_SSE2"
{
/* This insn should be already split before reg-stack. */
gcc_unreachable ();
}
[(set_attr "type" "multi")
(set_attr "unit" "sse,*,*")
(set_attr "mode" "TF,SI,SI")])
(define_split
[(set (match_operand:TF 0 "push_operand" "")
(match_operand:TF 1 "general_operand" ""))]
"TARGET_SSE2 && reload_completed
&& !SSE_REG_P (operands[1])"
[(const_int 0)]
"ix86_split_long_move (operands); DONE;")
(define_split
[(set (match_operand:TF 0 "push_operand" "")
(match_operand:TF 1 "any_fp_register_operand" ""))]
"TARGET_SSE2"
[(set (reg:P SP_REG) (plus:P (reg:P SP_REG) (const_int -16)))
(set (mem:TF (reg:P SP_REG)) (match_dup 1))]
"")
(define_split
[(set (match_operand 0 "nonimmediate_operand" "")
(match_operand 1 "general_operand" ""))]
"reload_completed
&& !(MEM_P (operands[0]) && MEM_P (operands[1]))
&& GET_MODE (operands[0]) == XFmode
&& ! (ANY_FP_REG_P (operands[0]) ||
(GET_CODE (operands[0]) == SUBREG
&& ANY_FP_REG_P (SUBREG_REG (operands[0]))))
&& ! (ANY_FP_REG_P (operands[1]) ||
(GET_CODE (operands[1]) == SUBREG
&& ANY_FP_REG_P (SUBREG_REG (operands[1]))))"
[(const_int 0)]
"ix86_split_long_move (operands); DONE;")
(define_split
[(set (match_operand 0 "register_operand" "")
(match_operand 1 "memory_operand" ""))]
"reload_completed
&& MEM_P (operands[1])
&& (GET_MODE (operands[0]) == TFmode
|| GET_MODE (operands[0]) == XFmode
|| GET_MODE (operands[0]) == SFmode
|| GET_MODE (operands[0]) == DFmode)
&& (operands[2] = find_constant_src (insn))"
[(set (match_dup 0) (match_dup 2))]
{
rtx c = operands[2];
rtx r = operands[0];
if (GET_CODE (r) == SUBREG)
r = SUBREG_REG (r);
if (SSE_REG_P (r))
{
if (!standard_sse_constant_p (c))
FAIL;
}
else if (FP_REG_P (r))
{
if (!standard_80387_constant_p (c))
FAIL;
}
else if (MMX_REG_P (r))
FAIL;
})
(define_split
[(set (match_operand 0 "register_operand" "")
(float_extend (match_operand 1 "memory_operand" "")))]
"reload_completed
&& MEM_P (operands[1])
&& (GET_MODE (operands[0]) == TFmode
|| GET_MODE (operands[0]) == XFmode
|| GET_MODE (operands[0]) == SFmode
|| GET_MODE (operands[0]) == DFmode)
&& (operands[2] = find_constant_src (insn))"
[(set (match_dup 0) (match_dup 2))]
{
rtx c = operands[2];
rtx r = operands[0];
if (GET_CODE (r) == SUBREG)
r = SUBREG_REG (r);
if (SSE_REG_P (r))
{
if (!standard_sse_constant_p (c))
FAIL;
}
else if (FP_REG_P (r))
{
if (!standard_80387_constant_p (c))
FAIL;
}
else if (MMX_REG_P (r))
FAIL;
})
(define_insn "swapxf"
[(set (match_operand:XF 0 "register_operand" "+f")
(match_operand:XF 1 "register_operand" "+f"))
(set (match_dup 1)
(match_dup 0))]
"TARGET_80387"
{
if (STACK_TOP_P (operands[0]))
return "fxch\t%1";
else
return "fxch\t%0";
}
[(set_attr "type" "fxch")
(set_attr "mode" "XF")])
;; Split the load of -0.0 or -1.0 into fldz;fchs or fld1;fchs sequence
(define_split
[(set (match_operand:X87MODEF 0 "register_operand" "")
(match_operand:X87MODEF 1 "immediate_operand" ""))]
"reload_completed && FP_REGNO_P (REGNO (operands[0]))
&& (standard_80387_constant_p (operands[1]) == 8
|| standard_80387_constant_p (operands[1]) == 9)"
[(set (match_dup 0)(match_dup 1))
(set (match_dup 0)
(neg:X87MODEF (match_dup 0)))]
{
REAL_VALUE_TYPE r;
REAL_VALUE_FROM_CONST_DOUBLE (r, operands[1]);
if (real_isnegzero (&r))
operands[1] = CONST0_RTX (mode);
else
operands[1] = CONST1_RTX (mode);
})
(define_split
[(set (match_operand:TF 0 "nonimmediate_operand" "")
(match_operand:TF 1 "general_operand" ""))]
"reload_completed
&& !(SSE_REG_P (operands[0]) || SSE_REG_P (operands[1]))"
[(const_int 0)]
"ix86_split_long_move (operands); DONE;")
;; Zero extension instructions
(define_expand "zero_extendhisi2"
[(set (match_operand:SI 0 "register_operand" "")
(zero_extend:SI (match_operand:HI 1 "nonimmediate_operand" "")))]
""
{
if (TARGET_ZERO_EXTEND_WITH_AND && optimize_function_for_speed_p (cfun))
{
operands[1] = force_reg (HImode, operands[1]);
emit_insn (gen_zero_extendhisi2_and (operands[0], operands[1]));
DONE;
}
})
(define_insn "zero_extendhisi2_and"
[(set (match_operand:SI 0 "register_operand" "=r")
(zero_extend:SI (match_operand:HI 1 "register_operand" "0")))
(clobber (reg:CC FLAGS_REG))]
"TARGET_ZERO_EXTEND_WITH_AND && optimize_function_for_speed_p (cfun)"
"#"
[(set_attr "type" "alu1")
(set_attr "mode" "SI")])
(define_split
[(set (match_operand:SI 0 "register_operand" "")
(zero_extend:SI (match_operand:HI 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))]
"reload_completed && TARGET_ZERO_EXTEND_WITH_AND
&& optimize_function_for_speed_p (cfun)"
[(parallel [(set (match_dup 0) (and:SI (match_dup 0) (const_int 65535)))
(clobber (reg:CC FLAGS_REG))])]
"")
(define_insn "*zero_extendhisi2_movzwl"
[(set (match_operand:SI 0 "register_operand" "=r")
(zero_extend:SI (match_operand:HI 1 "nonimmediate_operand" "rm")))]
"!TARGET_ZERO_EXTEND_WITH_AND
|| optimize_function_for_size_p (cfun)"
"movz{wl|x}\t{%1, %0|%0, %1}"
[(set_attr "type" "imovx")
(set_attr "mode" "SI")])
(define_expand "zero_extendqihi2"
[(parallel
[(set (match_operand:HI 0 "register_operand" "")
(zero_extend:HI (match_operand:QI 1 "nonimmediate_operand" "")))
(clobber (reg:CC FLAGS_REG))])]
""
"")
(define_insn "*zero_extendqihi2_and"
[(set (match_operand:HI 0 "register_operand" "=r,?&q")
(zero_extend:HI (match_operand:QI 1 "nonimmediate_operand" "0,qm")))
(clobber (reg:CC FLAGS_REG))]
"TARGET_ZERO_EXTEND_WITH_AND && optimize_function_for_speed_p (cfun)"
"#"
[(set_attr "type" "alu1")
(set_attr "mode" "HI")])
(define_insn "*zero_extendqihi2_movzbw_and"
[(set (match_operand:HI 0 "register_operand" "=r,r")
(zero_extend:HI (match_operand:QI 1 "nonimmediate_operand" "qm,0")))
(clobber (reg:CC FLAGS_REG))]
"!TARGET_ZERO_EXTEND_WITH_AND || optimize_function_for_size_p (cfun)"
"#"
[(set_attr "type" "imovx,alu1")
(set_attr "mode" "HI")])
; zero extend to SImode here to avoid partial register stalls
(define_insn "*zero_extendqihi2_movzbl"
[(set (match_operand:HI 0 "register_operand" "=r")
(zero_extend:HI (match_operand:QI 1 "nonimmediate_operand" "qm")))]
"(!TARGET_ZERO_EXTEND_WITH_AND || optimize_function_for_size_p (cfun))
&& reload_completed"
"movz{bl|x}\t{%1, %k0|%k0, %1}"
[(set_attr "type" "imovx")
(set_attr "mode" "SI")])
;; For the movzbw case strip only the clobber
(define_split
[(set (match_operand:HI 0 "register_operand" "")
(zero_extend:HI (match_operand:QI 1 "nonimmediate_operand" "")))
(clobber (reg:CC FLAGS_REG))]
"reload_completed
&& (!TARGET_ZERO_EXTEND_WITH_AND
|| optimize_function_for_size_p (cfun))
&& (!REG_P (operands[1]) || ANY_QI_REG_P (operands[1]))"
[(set (match_operand:HI 0 "register_operand" "")
(zero_extend:HI (match_operand:QI 1 "nonimmediate_operand" "")))])
;; When source and destination does not overlap, clear destination
;; first and then do the movb
(define_split
[(set (match_operand:HI 0 "register_operand" "")
(zero_extend:HI (match_operand:QI 1 "nonimmediate_operand" "")))
(clobber (reg:CC FLAGS_REG))]
"reload_completed
&& ANY_QI_REG_P (operands[0])
&& (TARGET_ZERO_EXTEND_WITH_AND
&& optimize_function_for_speed_p (cfun))
&& !reg_overlap_mentioned_p (operands[0], operands[1])"
[(set (match_dup 0) (const_int 0))
(set (strict_low_part (match_dup 2)) (match_dup 1))]
"operands[2] = gen_lowpart (QImode, operands[0]);")
;; Rest is handled by single and.
(define_split
[(set (match_operand:HI 0 "register_operand" "")
(zero_extend:HI (match_operand:QI 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))]
"reload_completed
&& true_regnum (operands[0]) == true_regnum (operands[1])"
[(parallel [(set (match_dup 0) (and:HI (match_dup 0) (const_int 255)))
(clobber (reg:CC FLAGS_REG))])]
"")
(define_expand "zero_extendqisi2"
[(parallel
[(set (match_operand:SI 0 "register_operand" "")
(zero_extend:SI (match_operand:QI 1 "nonimmediate_operand" "")))
(clobber (reg:CC FLAGS_REG))])]
""
"")
(define_insn "*zero_extendqisi2_and"
[(set (match_operand:SI 0 "register_operand" "=r,?&q")
(zero_extend:SI (match_operand:QI 1 "nonimmediate_operand" "0,qm")))
(clobber (reg:CC FLAGS_REG))]
"TARGET_ZERO_EXTEND_WITH_AND && optimize_function_for_speed_p (cfun)"
"#"
[(set_attr "type" "alu1")
(set_attr "mode" "SI")])
(define_insn "*zero_extendqisi2_movzbw_and"
[(set (match_operand:SI 0 "register_operand" "=r,r")
(zero_extend:SI (match_operand:QI 1 "nonimmediate_operand" "qm,0")))
(clobber (reg:CC FLAGS_REG))]
"!TARGET_ZERO_EXTEND_WITH_AND || optimize_function_for_size_p (cfun)"
"#"
[(set_attr "type" "imovx,alu1")
(set_attr "mode" "SI")])
(define_insn "*zero_extendqisi2_movzbw"
[(set (match_operand:SI 0 "register_operand" "=r")
(zero_extend:SI (match_operand:QI 1 "nonimmediate_operand" "qm")))]
"(!TARGET_ZERO_EXTEND_WITH_AND || optimize_function_for_size_p (cfun))
&& reload_completed"
"movz{bl|x}\t{%1, %0|%0, %1}"
[(set_attr "type" "imovx")
(set_attr "mode" "SI")])
;; For the movzbl case strip only the clobber
(define_split
[(set (match_operand:SI 0 "register_operand" "")
(zero_extend:SI (match_operand:QI 1 "nonimmediate_operand" "")))
(clobber (reg:CC FLAGS_REG))]
"reload_completed
&& (!TARGET_ZERO_EXTEND_WITH_AND || optimize_function_for_size_p (cfun))
&& (!REG_P (operands[1]) || ANY_QI_REG_P (operands[1]))"
[(set (match_dup 0)
(zero_extend:SI (match_dup 1)))])
;; When source and destination does not overlap, clear destination
;; first and then do the movb
(define_split
[(set (match_operand:SI 0 "register_operand" "")
(zero_extend:SI (match_operand:QI 1 "nonimmediate_operand" "")))
(clobber (reg:CC FLAGS_REG))]
"reload_completed
&& ANY_QI_REG_P (operands[0])
&& (ANY_QI_REG_P (operands[1]) || MEM_P (operands[1]))
&& (TARGET_ZERO_EXTEND_WITH_AND && optimize_function_for_speed_p (cfun))
&& !reg_overlap_mentioned_p (operands[0], operands[1])"
[(set (match_dup 0) (const_int 0))
(set (strict_low_part (match_dup 2)) (match_dup 1))]
"operands[2] = gen_lowpart (QImode, operands[0]);")
;; Rest is handled by single and.
(define_split
[(set (match_operand:SI 0 "register_operand" "")
(zero_extend:SI (match_operand:QI 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))]
"reload_completed
&& true_regnum (operands[0]) == true_regnum (operands[1])"
[(parallel [(set (match_dup 0) (and:SI (match_dup 0) (const_int 255)))
(clobber (reg:CC FLAGS_REG))])]
"")
;; %%% Kill me once multi-word ops are sane.
(define_expand "zero_extendsidi2"
[(set (match_operand:DI 0 "register_operand" "")
(zero_extend:DI (match_operand:SI 1 "nonimmediate_operand" "")))]
""
{
if (!TARGET_64BIT)
{
emit_insn (gen_zero_extendsidi2_32 (operands[0], operands[1]));
DONE;
}
})
(define_insn "zero_extendsidi2_32"
[(set (match_operand:DI 0 "nonimmediate_operand" "=r,?r,?o,?*Ym,?*y,?*Yi,*Y2")
(zero_extend:DI
(match_operand:SI 1 "nonimmediate_operand" "0,rm,r ,r ,m ,r ,m")))
(clobber (reg:CC FLAGS_REG))]
"!TARGET_64BIT"
"@
#
#
#
movd\t{%1, %0|%0, %1}
movd\t{%1, %0|%0, %1}
%vmovd\t{%1, %0|%0, %1}
%vmovd\t{%1, %0|%0, %1}"
[(set_attr "type" "multi,multi,multi,mmxmov,mmxmov,ssemov,ssemov")
(set_attr "prefix" "*,*,*,orig,orig,maybe_vex,maybe_vex")
(set_attr "mode" "SI,SI,SI,DI,DI,TI,TI")])
(define_insn "zero_extendsidi2_rex64"
[(set (match_operand:DI 0 "nonimmediate_operand" "=r,o,?*Ym,?*y,?*Yi,*Y2")
(zero_extend:DI
(match_operand:SI 1 "nonimmediate_operand" "rm,0,r ,m ,r ,m")))]
"TARGET_64BIT"
"@
mov\t{%k1, %k0|%k0, %k1}
#
movd\t{%1, %0|%0, %1}
movd\t{%1, %0|%0, %1}
%vmovd\t{%1, %0|%0, %1}
%vmovd\t{%1, %0|%0, %1}"
[(set_attr "type" "imovx,imov,mmxmov,mmxmov,ssemov,ssemov")
(set_attr "prefix" "orig,*,orig,orig,maybe_vex,maybe_vex")
(set_attr "prefix_0f" "0,*,*,*,*,*")
(set_attr "mode" "SI,DI,DI,DI,TI,TI")])
(define_split
[(set (match_operand:DI 0 "memory_operand" "")
(zero_extend:DI (match_dup 0)))]
"TARGET_64BIT"
[(set (match_dup 4) (const_int 0))]
"split_di (&operands[0], 1, &operands[3], &operands[4]);")
(define_split
[(set (match_operand:DI 0 "register_operand" "")
(zero_extend:DI (match_operand:SI 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))]
"!TARGET_64BIT && reload_completed
&& true_regnum (operands[0]) == true_regnum (operands[1])"
[(set (match_dup 4) (const_int 0))]
"split_di (&operands[0], 1, &operands[3], &operands[4]);")
(define_split
[(set (match_operand:DI 0 "nonimmediate_operand" "")
(zero_extend:DI (match_operand:SI 1 "general_operand" "")))
(clobber (reg:CC FLAGS_REG))]
"!TARGET_64BIT && reload_completed
&& !SSE_REG_P (operands[0]) && !MMX_REG_P (operands[0])"
[(set (match_dup 3) (match_dup 1))
(set (match_dup 4) (const_int 0))]
"split_di (&operands[0], 1, &operands[3], &operands[4]);")
(define_insn "zero_extendhidi2"
[(set (match_operand:DI 0 "register_operand" "=r")
(zero_extend:DI (match_operand:HI 1 "nonimmediate_operand" "rm")))]
"TARGET_64BIT"
"movz{wl|x}\t{%1, %k0|%k0, %1}"
[(set_attr "type" "imovx")
(set_attr "mode" "SI")])
(define_insn "zero_extendqidi2"
[(set (match_operand:DI 0 "register_operand" "=r")
(zero_extend:DI (match_operand:QI 1 "nonimmediate_operand" "rm")))]
"TARGET_64BIT"
"movz{bl|x}\t{%1, %k0|%k0, %1}"
[(set_attr "type" "imovx")
(set_attr "mode" "SI")])
;; Sign extension instructions
(define_expand "extendsidi2"
[(parallel [(set (match_operand:DI 0 "register_operand" "")
(sign_extend:DI (match_operand:SI 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))
(clobber (match_scratch:SI 2 ""))])]
""
{
if (TARGET_64BIT)
{
emit_insn (gen_extendsidi2_rex64 (operands[0], operands[1]));
DONE;
}
})
(define_insn "*extendsidi2_1"
[(set (match_operand:DI 0 "nonimmediate_operand" "=*A,r,?r,?*o")
(sign_extend:DI (match_operand:SI 1 "register_operand" "0,0,r,r")))
(clobber (reg:CC FLAGS_REG))
(clobber (match_scratch:SI 2 "=X,X,X,&r"))]
"!TARGET_64BIT"
"#")
(define_insn "extendsidi2_rex64"
[(set (match_operand:DI 0 "register_operand" "=*a,r")
(sign_extend:DI (match_operand:SI 1 "nonimmediate_operand" "*0,rm")))]
"TARGET_64BIT"
"@
{cltq|cdqe}
movs{lq|x}\t{%1, %0|%0, %1}"
[(set_attr "type" "imovx")
(set_attr "mode" "DI")
(set_attr "prefix_0f" "0")
(set_attr "modrm" "0,1")])
(define_insn "extendhidi2"
[(set (match_operand:DI 0 "register_operand" "=r")
(sign_extend:DI (match_operand:HI 1 "nonimmediate_operand" "rm")))]
"TARGET_64BIT"
"movs{wq|x}\t{%1, %0|%0, %1}"
[(set_attr "type" "imovx")
(set_attr "mode" "DI")])
(define_insn "extendqidi2"
[(set (match_operand:DI 0 "register_operand" "=r")
(sign_extend:DI (match_operand:QI 1 "nonimmediate_operand" "qm")))]
"TARGET_64BIT"
"movs{bq|x}\t{%1, %0|%0, %1}"
[(set_attr "type" "imovx")
(set_attr "mode" "DI")])
;; Extend to memory case when source register does die.
(define_split
[(set (match_operand:DI 0 "memory_operand" "")
(sign_extend:DI (match_operand:SI 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))
(clobber (match_operand:SI 2 "register_operand" ""))]
"(reload_completed
&& dead_or_set_p (insn, operands[1])
&& !reg_mentioned_p (operands[1], operands[0]))"
[(set (match_dup 3) (match_dup 1))
(parallel [(set (match_dup 1) (ashiftrt:SI (match_dup 1) (const_int 31)))
(clobber (reg:CC FLAGS_REG))])
(set (match_dup 4) (match_dup 1))]
"split_di (&operands[0], 1, &operands[3], &operands[4]);")
;; Extend to memory case when source register does not die.
(define_split
[(set (match_operand:DI 0 "memory_operand" "")
(sign_extend:DI (match_operand:SI 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))
(clobber (match_operand:SI 2 "register_operand" ""))]
"reload_completed"
[(const_int 0)]
{
split_di (&operands[0], 1, &operands[3], &operands[4]);
emit_move_insn (operands[3], operands[1]);
/* Generate a cltd if possible and doing so it profitable. */
if ((optimize_function_for_size_p (cfun) || TARGET_USE_CLTD)
&& true_regnum (operands[1]) == AX_REG
&& true_regnum (operands[2]) == DX_REG)
{
emit_insn (gen_ashrsi3_31 (operands[2], operands[1], GEN_INT (31)));
}
else
{
emit_move_insn (operands[2], operands[1]);
emit_insn (gen_ashrsi3_31 (operands[2], operands[2], GEN_INT (31)));
}
emit_move_insn (operands[4], operands[2]);
DONE;
})
;; Extend to register case. Optimize case where source and destination
;; registers match and cases where we can use cltd.
(define_split
[(set (match_operand:DI 0 "register_operand" "")
(sign_extend:DI (match_operand:SI 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))
(clobber (match_scratch:SI 2 ""))]
"reload_completed"
[(const_int 0)]
{
split_di (&operands[0], 1, &operands[3], &operands[4]);
if (true_regnum (operands[3]) != true_regnum (operands[1]))
emit_move_insn (operands[3], operands[1]);
/* Generate a cltd if possible and doing so it profitable. */
if ((optimize_function_for_size_p (cfun) || TARGET_USE_CLTD)
&& true_regnum (operands[3]) == AX_REG)
{
emit_insn (gen_ashrsi3_31 (operands[4], operands[3], GEN_INT (31)));
DONE;
}
if (true_regnum (operands[4]) != true_regnum (operands[1]))
emit_move_insn (operands[4], operands[1]);
emit_insn (gen_ashrsi3_31 (operands[4], operands[4], GEN_INT (31)));
DONE;
})
(define_insn "extendhisi2"
[(set (match_operand:SI 0 "register_operand" "=*a,r")
(sign_extend:SI (match_operand:HI 1 "nonimmediate_operand" "*0,rm")))]
""
{
switch (get_attr_prefix_0f (insn))
{
case 0:
return "{cwtl|cwde}";
default:
return "movs{wl|x}\t{%1, %0|%0, %1}";
}
}
[(set_attr "type" "imovx")
(set_attr "mode" "SI")
(set (attr "prefix_0f")
;; movsx is short decodable while cwtl is vector decoded.
(if_then_else (and (eq_attr "cpu" "!k6")
(eq_attr "alternative" "0"))
(const_string "0")
(const_string "1")))
(set (attr "modrm")
(if_then_else (eq_attr "prefix_0f" "0")
(const_string "0")
(const_string "1")))])
(define_insn "*extendhisi2_zext"
[(set (match_operand:DI 0 "register_operand" "=*a,r")
(zero_extend:DI
(sign_extend:SI (match_operand:HI 1 "nonimmediate_operand" "*0,rm"))))]
"TARGET_64BIT"
{
switch (get_attr_prefix_0f (insn))
{
case 0:
return "{cwtl|cwde}";
default:
return "movs{wl|x}\t{%1, %k0|%k0, %1}";
}
}
[(set_attr "type" "imovx")
(set_attr "mode" "SI")
(set (attr "prefix_0f")
;; movsx is short decodable while cwtl is vector decoded.
(if_then_else (and (eq_attr "cpu" "!k6")
(eq_attr "alternative" "0"))
(const_string "0")
(const_string "1")))
(set (attr "modrm")
(if_then_else (eq_attr "prefix_0f" "0")
(const_string "0")
(const_string "1")))])
(define_insn "extendqihi2"
[(set (match_operand:HI 0 "register_operand" "=*a,r")
(sign_extend:HI (match_operand:QI 1 "nonimmediate_operand" "*0,qm")))]
""
{
switch (get_attr_prefix_0f (insn))
{
case 0:
return "{cbtw|cbw}";
default:
return "movs{bw|x}\t{%1, %0|%0, %1}";
}
}
[(set_attr "type" "imovx")
(set_attr "mode" "HI")
(set (attr "prefix_0f")
;; movsx is short decodable while cwtl is vector decoded.
(if_then_else (and (eq_attr "cpu" "!k6")
(eq_attr "alternative" "0"))
(const_string "0")
(const_string "1")))
(set (attr "modrm")
(if_then_else (eq_attr "prefix_0f" "0")
(const_string "0")
(const_string "1")))])
(define_insn "extendqisi2"
[(set (match_operand:SI 0 "register_operand" "=r")
(sign_extend:SI (match_operand:QI 1 "nonimmediate_operand" "qm")))]
""
"movs{bl|x}\t{%1, %0|%0, %1}"
[(set_attr "type" "imovx")
(set_attr "mode" "SI")])
(define_insn "*extendqisi2_zext"
[(set (match_operand:DI 0 "register_operand" "=r")
(zero_extend:DI
(sign_extend:SI (match_operand:QI 1 "nonimmediate_operand" "qm"))))]
"TARGET_64BIT"
"movs{bl|x}\t{%1, %k0|%k0, %1}"
[(set_attr "type" "imovx")
(set_attr "mode" "SI")])
;; Conversions between float and double.
;; These are all no-ops in the model used for the 80387. So just
;; emit moves.
;; %%% Kill these when call knows how to work out a DFmode push earlier.
(define_insn "*dummy_extendsfdf2"
[(set (match_operand:DF 0 "push_operand" "=<")
(float_extend:DF (match_operand:SF 1 "nonimmediate_operand" "fY2")))]
"0"
"#")
(define_split
[(set (match_operand:DF 0 "push_operand" "")
(float_extend:DF (match_operand:SF 1 "fp_register_operand" "")))]
""
[(set (reg:P SP_REG) (plus:P (reg:P SP_REG) (const_int -8)))
(set (mem:DF (reg:P SP_REG)) (float_extend:DF (match_dup 1)))])
(define_insn "*dummy_extendsfxf2"
[(set (match_operand:XF 0 "push_operand" "=<")
(float_extend:XF (match_operand:SF 1 "nonimmediate_operand" "f")))]
"0"
"#")
(define_split
[(set (match_operand:XF 0 "push_operand" "")
(float_extend:XF (match_operand:SF 1 "fp_register_operand" "")))]
""
[(set (reg:P SP_REG) (plus:P (reg:P SP_REG) (match_dup 2)))
(set (mem:XF (reg:P SP_REG)) (float_extend:XF (match_dup 1)))]
"operands[2] = GEN_INT (TARGET_128BIT_LONG_DOUBLE ? -16 : -12);")
(define_split
[(set (match_operand:XF 0 "push_operand" "")
(float_extend:XF (match_operand:DF 1 "fp_register_operand" "")))]
""
[(set (reg:P SP_REG) (plus:P (reg:P SP_REG) (match_dup 2)))
(set (mem:DF (reg:P SP_REG)) (float_extend:XF (match_dup 1)))]
"operands[2] = GEN_INT (TARGET_128BIT_LONG_DOUBLE ? -16 : -12);")
(define_expand "extendsfdf2"
[(set (match_operand:DF 0 "nonimmediate_operand" "")
(float_extend:DF (match_operand:SF 1 "general_operand" "")))]
"TARGET_80387 || (TARGET_SSE2 && TARGET_SSE_MATH)"
{
/* ??? Needed for compress_float_constant since all fp constants
are LEGITIMATE_CONSTANT_P. */
if (GET_CODE (operands[1]) == CONST_DOUBLE)
{
if ((!TARGET_SSE2 || TARGET_MIX_SSE_I387)
&& standard_80387_constant_p (operands[1]) > 0)
{
operands[1] = simplify_const_unary_operation
(FLOAT_EXTEND, DFmode, operands[1], SFmode);
emit_move_insn_1 (operands[0], operands[1]);
DONE;
}
operands[1] = validize_mem (force_const_mem (SFmode, operands[1]));
}
})
/* For converting SF(xmm2) to DF(xmm1), use the following code instead of
cvtss2sd:
unpcklps xmm2,xmm2 ; packed conversion might crash on signaling NaNs
cvtps2pd xmm2,xmm1
We do the conversion post reload to avoid producing of 128bit spills
that might lead to ICE on 32bit target. The sequence unlikely combine
anyway. */
(define_split
[(set (match_operand:DF 0 "register_operand" "")
(float_extend:DF
(match_operand:SF 1 "nonimmediate_operand" "")))]
"TARGET_USE_VECTOR_FP_CONVERTS
&& optimize_insn_for_speed_p ()
&& reload_completed && SSE_REG_P (operands[0])"
[(set (match_dup 2)
(float_extend:V2DF
(vec_select:V2SF
(match_dup 3)
(parallel [(const_int 0) (const_int 1)]))))]
{
operands[2] = simplify_gen_subreg (V2DFmode, operands[0], DFmode, 0);
operands[3] = simplify_gen_subreg (V4SFmode, operands[0], DFmode, 0);
/* Use movss for loading from memory, unpcklps reg, reg for registers.
Try to avoid move when unpacking can be done in source. */
if (REG_P (operands[1]))
{
/* If it is unsafe to overwrite upper half of source, we need
to move to destination and unpack there. */
if ((ORIGINAL_REGNO (operands[1]) < FIRST_PSEUDO_REGISTER
|| PSEUDO_REGNO_BYTES (ORIGINAL_REGNO (operands[1])) > 4)
&& true_regnum (operands[0]) != true_regnum (operands[1]))
{
rtx tmp = gen_rtx_REG (SFmode, true_regnum (operands[0]));
emit_move_insn (tmp, operands[1]);
}
else
operands[3] = simplify_gen_subreg (V4SFmode, operands[1], SFmode, 0);
emit_insn (gen_sse_unpcklps (operands[3], operands[3], operands[3]));
}
else
emit_insn (gen_vec_setv4sf_0 (operands[3],
CONST0_RTX (V4SFmode), operands[1]));
})
(define_insn "*extendsfdf2_mixed"
[(set (match_operand:DF 0 "nonimmediate_operand" "=f,m,x")
(float_extend:DF
(match_operand:SF 1 "nonimmediate_operand" "fm,f,xm")))]
"TARGET_SSE2 && TARGET_MIX_SSE_I387"
{
switch (which_alternative)
{
case 0:
case 1:
return output_387_reg_move (insn, operands);
case 2:
return "%vcvtss2sd\t{%1, %d0|%d0, %1}";
default:
gcc_unreachable ();
}
}
[(set_attr "type" "fmov,fmov,ssecvt")
(set_attr "prefix" "orig,orig,maybe_vex")
(set_attr "mode" "SF,XF,DF")])
(define_insn "*extendsfdf2_sse"
[(set (match_operand:DF 0 "nonimmediate_operand" "=x")
(float_extend:DF (match_operand:SF 1 "nonimmediate_operand" "xm")))]
"TARGET_SSE2 && TARGET_SSE_MATH"
"%vcvtss2sd\t{%1, %d0|%d0, %1}"
[(set_attr "type" "ssecvt")
(set_attr "prefix" "maybe_vex")
(set_attr "mode" "DF")])
(define_insn "*extendsfdf2_i387"
[(set (match_operand:DF 0 "nonimmediate_operand" "=f,m")
(float_extend:DF (match_operand:SF 1 "nonimmediate_operand" "fm,f")))]
"TARGET_80387"
"* return output_387_reg_move (insn, operands);"
[(set_attr "type" "fmov")
(set_attr "mode" "SF,XF")])
(define_expand "extendxf2"
[(set (match_operand:XF 0 "nonimmediate_operand" "")
(float_extend:XF (match_operand:MODEF 1 "general_operand" "")))]
"TARGET_80387"
{
/* ??? Needed for compress_float_constant since all fp constants
are LEGITIMATE_CONSTANT_P. */
if (GET_CODE (operands[1]) == CONST_DOUBLE)
{
if (standard_80387_constant_p (operands[1]) > 0)
{
operands[1] = simplify_const_unary_operation
(FLOAT_EXTEND, XFmode, operands[1], mode);
emit_move_insn_1 (operands[0], operands[1]);
DONE;
}
operands[1] = validize_mem (force_const_mem (mode, operands[1]));
}
})
(define_insn "*extendxf2_i387"
[(set (match_operand:XF 0 "nonimmediate_operand" "=f,m")
(float_extend:XF
(match_operand:MODEF 1 "nonimmediate_operand" "fm,f")))]
"TARGET_80387"
"* return output_387_reg_move (insn, operands);"
[(set_attr "type" "fmov")
(set_attr "mode" ",XF")])
;; %%% This seems bad bad news.
;; This cannot output into an f-reg because there is no way to be sure
;; of truncating in that case. Otherwise this is just like a simple move
;; insn. So we pretend we can output to a reg in order to get better
;; register preferencing, but we really use a stack slot.
;; Conversion from DFmode to SFmode.
(define_expand "truncdfsf2"
[(set (match_operand:SF 0 "nonimmediate_operand" "")
(float_truncate:SF
(match_operand:DF 1 "nonimmediate_operand" "")))]
"TARGET_80387 || (TARGET_SSE2 && TARGET_SSE_MATH)"
{
if (TARGET_SSE2 && TARGET_SSE_MATH && !TARGET_MIX_SSE_I387)
;
else if (flag_unsafe_math_optimizations)
;
else
{
enum ix86_stack_slot slot = (virtuals_instantiated
? SLOT_TEMP
: SLOT_VIRTUAL);
rtx temp = assign_386_stack_local (SFmode, slot);
emit_insn (gen_truncdfsf2_with_temp (operands[0], operands[1], temp));
DONE;
}
})
/* For converting DF(xmm2) to SF(xmm1), use the following code instead of
cvtsd2ss:
unpcklpd xmm2,xmm2 ; packed conversion might crash on signaling NaNs
cvtpd2ps xmm2,xmm1
We do the conversion post reload to avoid producing of 128bit spills
that might lead to ICE on 32bit target. The sequence unlikely combine
anyway. */
(define_split
[(set (match_operand:SF 0 "register_operand" "")
(float_truncate:SF
(match_operand:DF 1 "nonimmediate_operand" "")))]
"TARGET_USE_VECTOR_FP_CONVERTS
&& optimize_insn_for_speed_p ()
&& reload_completed && SSE_REG_P (operands[0])"
[(set (match_dup 2)
(vec_concat:V4SF
(float_truncate:V2SF
(match_dup 4))
(match_dup 3)))]
{
operands[2] = simplify_gen_subreg (V4SFmode, operands[0], SFmode, 0);
operands[3] = CONST0_RTX (V2SFmode);
operands[4] = simplify_gen_subreg (V2DFmode, operands[0], SFmode, 0);
/* Use movsd for loading from memory, unpcklpd for registers.
Try to avoid move when unpacking can be done in source, or SSE3
movddup is available. */
if (REG_P (operands[1]))
{
if (!TARGET_SSE3
&& true_regnum (operands[0]) != true_regnum (operands[1])
&& (ORIGINAL_REGNO (operands[1]) < FIRST_PSEUDO_REGISTER
|| PSEUDO_REGNO_BYTES (ORIGINAL_REGNO (operands[1])) > 8))
{
rtx tmp = simplify_gen_subreg (DFmode, operands[0], SFmode, 0);
emit_move_insn (tmp, operands[1]);
operands[1] = tmp;
}
else if (!TARGET_SSE3)
operands[4] = simplify_gen_subreg (V2DFmode, operands[1], DFmode, 0);
emit_insn (gen_vec_dupv2df (operands[4], operands[1]));
}
else
emit_insn (gen_sse2_loadlpd (operands[4],
CONST0_RTX (V2DFmode), operands[1]));
})
(define_expand "truncdfsf2_with_temp"
[(parallel [(set (match_operand:SF 0 "" "")
(float_truncate:SF (match_operand:DF 1 "" "")))
(clobber (match_operand:SF 2 "" ""))])]
"")
(define_insn "*truncdfsf_fast_mixed"
[(set (match_operand:SF 0 "nonimmediate_operand" "=fm,x")
(float_truncate:SF
(match_operand:DF 1 "nonimmediate_operand" "f ,xm")))]
"TARGET_SSE2 && TARGET_MIX_SSE_I387 && flag_unsafe_math_optimizations"
{
switch (which_alternative)
{
case 0:
return output_387_reg_move (insn, operands);
case 1:
return "%vcvtsd2ss\t{%1, %d0|%d0, %1}";
default:
gcc_unreachable ();
}
}
[(set_attr "type" "fmov,ssecvt")
(set_attr "prefix" "orig,maybe_vex")
(set_attr "mode" "SF")])
;; Yes, this one doesn't depend on flag_unsafe_math_optimizations,
;; because nothing we do here is unsafe.
(define_insn "*truncdfsf_fast_sse"
[(set (match_operand:SF 0 "nonimmediate_operand" "=x")
(float_truncate:SF
(match_operand:DF 1 "nonimmediate_operand" "xm")))]
"TARGET_SSE2 && TARGET_SSE_MATH"
"%vcvtsd2ss\t{%1, %d0|%d0, %1}"
[(set_attr "type" "ssecvt")
(set_attr "prefix" "maybe_vex")
(set_attr "mode" "SF")])
(define_insn "*truncdfsf_fast_i387"
[(set (match_operand:SF 0 "nonimmediate_operand" "=fm")
(float_truncate:SF
(match_operand:DF 1 "nonimmediate_operand" "f")))]
"TARGET_80387 && flag_unsafe_math_optimizations"
"* return output_387_reg_move (insn, operands);"
[(set_attr "type" "fmov")
(set_attr "mode" "SF")])
(define_insn "*truncdfsf_mixed"
[(set (match_operand:SF 0 "nonimmediate_operand" "=m,Y2 ,?f,?x,?*r")
(float_truncate:SF
(match_operand:DF 1 "nonimmediate_operand" "f ,Y2m,f ,f ,f")))
(clobber (match_operand:SF 2 "memory_operand" "=X,X ,m ,m ,m"))]
"TARGET_MIX_SSE_I387"
{
switch (which_alternative)
{
case 0:
return output_387_reg_move (insn, operands);
case 1:
return "%vcvtsd2ss\t{%1, %d0|%d0, %1}";
default:
return "#";
}
}
[(set_attr "type" "fmov,ssecvt,multi,multi,multi")
(set_attr "unit" "*,*,i387,i387,i387")
(set_attr "prefix" "orig,maybe_vex,orig,orig,orig")
(set_attr "mode" "SF")])
(define_insn "*truncdfsf_i387"
[(set (match_operand:SF 0 "nonimmediate_operand" "=m,?f,?x,?*r")
(float_truncate:SF
(match_operand:DF 1 "nonimmediate_operand" "f ,f ,f ,f")))
(clobber (match_operand:SF 2 "memory_operand" "=X,m ,m ,m"))]
"TARGET_80387"
{
switch (which_alternative)
{
case 0:
return output_387_reg_move (insn, operands);
default:
return "#";
}
}
[(set_attr "type" "fmov,multi,multi,multi")
(set_attr "unit" "*,i387,i387,i387")
(set_attr "mode" "SF")])
(define_insn "*truncdfsf2_i387_1"
[(set (match_operand:SF 0 "memory_operand" "=m")
(float_truncate:SF
(match_operand:DF 1 "register_operand" "f")))]
"TARGET_80387
&& !(TARGET_SSE2 && TARGET_SSE_MATH)
&& !TARGET_MIX_SSE_I387"
"* return output_387_reg_move (insn, operands);"
[(set_attr "type" "fmov")
(set_attr "mode" "SF")])
(define_split
[(set (match_operand:SF 0 "register_operand" "")
(float_truncate:SF
(match_operand:DF 1 "fp_register_operand" "")))
(clobber (match_operand 2 "" ""))]
"reload_completed"
[(set (match_dup 2) (match_dup 1))
(set (match_dup 0) (match_dup 2))]
{
operands[1] = gen_rtx_REG (SFmode, true_regnum (operands[1]));
})
;; Conversion from XFmode to {SF,DF}mode
(define_expand "truncxf2"
[(parallel [(set (match_operand:MODEF 0 "nonimmediate_operand" "")
(float_truncate:MODEF
(match_operand:XF 1 "register_operand" "")))
(clobber (match_dup 2))])]
"TARGET_80387"
{
if (flag_unsafe_math_optimizations)
{
rtx reg = REG_P (operands[0]) ? operands[0] : gen_reg_rtx (mode);
emit_insn (gen_truncxf2_i387_noop (reg, operands[1]));
if (reg != operands[0])
emit_move_insn (operands[0], reg);
DONE;
}
else
{
enum ix86_stack_slot slot = (virtuals_instantiated
? SLOT_TEMP
: SLOT_VIRTUAL);
operands[2] = assign_386_stack_local (mode, slot);
}
})
(define_insn "*truncxfsf2_mixed"
[(set (match_operand:SF 0 "nonimmediate_operand" "=m,?f,?x,?*r")
(float_truncate:SF
(match_operand:XF 1 "register_operand" "f ,f ,f ,f")))
(clobber (match_operand:SF 2 "memory_operand" "=X,m ,m ,m"))]
"TARGET_80387"
{
gcc_assert (!which_alternative);
return output_387_reg_move (insn, operands);
}
[(set_attr "type" "fmov,multi,multi,multi")
(set_attr "unit" "*,i387,i387,i387")
(set_attr "mode" "SF")])
(define_insn "*truncxfdf2_mixed"
[(set (match_operand:DF 0 "nonimmediate_operand" "=m,?f,?Y2,?*r")
(float_truncate:DF
(match_operand:XF 1 "register_operand" "f ,f ,f ,f")))
(clobber (match_operand:DF 2 "memory_operand" "=X,m ,m ,m"))]
"TARGET_80387"
{
gcc_assert (!which_alternative);
return output_387_reg_move (insn, operands);
}
[(set_attr "type" "fmov,multi,multi,multi")
(set_attr "unit" "*,i387,i387,i387")
(set_attr "mode" "DF")])
(define_insn "truncxf2_i387_noop"
[(set (match_operand:MODEF 0 "register_operand" "=f")
(float_truncate:MODEF
(match_operand:XF 1 "register_operand" "f")))]
"TARGET_80387 && flag_unsafe_math_optimizations"
"* return output_387_reg_move (insn, operands);"
[(set_attr "type" "fmov")
(set_attr "mode" "")])
(define_insn "*truncxf2_i387"
[(set (match_operand:MODEF 0 "memory_operand" "=m")
(float_truncate:MODEF
(match_operand:XF 1 "register_operand" "f")))]
"TARGET_80387"
"* return output_387_reg_move (insn, operands);"
[(set_attr "type" "fmov")
(set_attr "mode" "")])
(define_split
[(set (match_operand:MODEF 0 "register_operand" "")
(float_truncate:MODEF
(match_operand:XF 1 "register_operand" "")))
(clobber (match_operand:MODEF 2 "memory_operand" ""))]
"TARGET_80387 && reload_completed"
[(set (match_dup 2) (float_truncate:MODEF (match_dup 1)))
(set (match_dup 0) (match_dup 2))]
"")
(define_split
[(set (match_operand:MODEF 0 "memory_operand" "")
(float_truncate:MODEF
(match_operand:XF 1 "register_operand" "")))
(clobber (match_operand:MODEF 2 "memory_operand" ""))]
"TARGET_80387"
[(set (match_dup 0) (float_truncate:MODEF (match_dup 1)))]
"")
;; Signed conversion to DImode.
(define_expand "fix_truncxfdi2"
[(parallel [(set (match_operand:DI 0 "nonimmediate_operand" "")
(fix:DI (match_operand:XF 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))])]
"TARGET_80387"
{
if (TARGET_FISTTP)
{
emit_insn (gen_fix_truncdi_fisttp_i387_1 (operands[0], operands[1]));
DONE;
}
})
(define_expand "fix_truncdi2"
[(parallel [(set (match_operand:DI 0 "nonimmediate_operand" "")
(fix:DI (match_operand:MODEF 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))])]
"TARGET_80387 || (TARGET_64BIT && SSE_FLOAT_MODE_P (mode))"
{
if (TARGET_FISTTP
&& !(TARGET_64BIT && SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH))
{
emit_insn (gen_fix_truncdi_fisttp_i387_1 (operands[0], operands[1]));
DONE;
}
if (TARGET_64BIT && SSE_FLOAT_MODE_P (mode))
{
rtx out = REG_P (operands[0]) ? operands[0] : gen_reg_rtx (DImode);
emit_insn (gen_fix_truncdi_sse (out, operands[1]));
if (out != operands[0])
emit_move_insn (operands[0], out);
DONE;
}
})
;; Signed conversion to SImode.
(define_expand "fix_truncxfsi2"
[(parallel [(set (match_operand:SI 0 "nonimmediate_operand" "")
(fix:SI (match_operand:XF 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))])]
"TARGET_80387"
{
if (TARGET_FISTTP)
{
emit_insn (gen_fix_truncsi_fisttp_i387_1 (operands[0], operands[1]));
DONE;
}
})
(define_expand "fix_truncsi2"
[(parallel [(set (match_operand:SI 0 "nonimmediate_operand" "")
(fix:SI (match_operand:MODEF 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))])]
"TARGET_80387 || SSE_FLOAT_MODE_P (mode)"
{
if (TARGET_FISTTP
&& !(SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH))
{
emit_insn (gen_fix_truncsi_fisttp_i387_1 (operands[0], operands[1]));
DONE;
}
if (SSE_FLOAT_MODE_P (mode))
{
rtx out = REG_P (operands[0]) ? operands[0] : gen_reg_rtx (SImode);
emit_insn (gen_fix_truncsi_sse (out, operands[1]));
if (out != operands[0])
emit_move_insn (operands[0], out);
DONE;
}
})
;; Signed conversion to HImode.
(define_expand "fix_trunchi2"
[(parallel [(set (match_operand:HI 0 "nonimmediate_operand" "")
(fix:HI (match_operand:X87MODEF 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))])]
"TARGET_80387
&& !(SSE_FLOAT_MODE_P (mode) && (!TARGET_FISTTP || TARGET_SSE_MATH))"
{
if (TARGET_FISTTP)
{
emit_insn (gen_fix_trunchi_fisttp_i387_1 (operands[0], operands[1]));
DONE;
}
})
;; Unsigned conversion to SImode.
(define_expand "fixuns_truncsi2"
[(parallel
[(set (match_operand:SI 0 "register_operand" "")
(unsigned_fix:SI
(match_operand:MODEF 1 "nonimmediate_operand" "")))
(use (match_dup 2))
(clobber (match_scratch: 3 ""))
(clobber (match_scratch: 4 ""))])]
"!TARGET_64BIT && TARGET_SSE2 && TARGET_SSE_MATH"
{
enum machine_mode mode = mode;
enum machine_mode vecmode = mode;
REAL_VALUE_TYPE TWO31r;
rtx two31;
if (optimize_insn_for_size_p ())
FAIL;
real_ldexp (&TWO31r, &dconst1, 31);
two31 = const_double_from_real_value (TWO31r, mode);
two31 = ix86_build_const_vector (mode, true, two31);
operands[2] = force_reg (vecmode, two31);
})
(define_insn_and_split "*fixuns_trunc_1"
[(set (match_operand:SI 0 "register_operand" "=&x,&x")
(unsigned_fix:SI
(match_operand:MODEF 3 "nonimmediate_operand" "xm,xm")))
(use (match_operand: 4 "nonimmediate_operand" "m,x"))
(clobber (match_scratch: 1 "=x,&x"))
(clobber (match_scratch: 2 "=x,x"))]
"!TARGET_64BIT && TARGET_SSE2 && TARGET_SSE_MATH
&& optimize_function_for_speed_p (cfun)"
"#"
"&& reload_completed"
[(const_int 0)]
{
ix86_split_convert_uns_si_sse (operands);
DONE;
})
;; Unsigned conversion to HImode.
;; Without these patterns, we'll try the unsigned SI conversion which
;; is complex for SSE, rather than the signed SI conversion, which isn't.
(define_expand "fixuns_trunchi2"
[(set (match_dup 2)
(fix:SI (match_operand:MODEF 1 "nonimmediate_operand" "")))
(set (match_operand:HI 0 "nonimmediate_operand" "")
(subreg:HI (match_dup 2) 0))]
"SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH"
"operands[2] = gen_reg_rtx (SImode);")
;; When SSE is available, it is always faster to use it!
(define_insn "fix_truncdi_sse"
[(set (match_operand:DI 0 "register_operand" "=r,r")
(fix:DI (match_operand:MODEF 1 "nonimmediate_operand" "x,m")))]
"TARGET_64BIT && SSE_FLOAT_MODE_P (mode)
&& (!TARGET_FISTTP || TARGET_SSE_MATH)"
"%vcvtts2si{q}\t{%1, %0|%0, %1}"
[(set_attr "type" "sseicvt")
(set_attr "prefix" "maybe_vex")
(set_attr "prefix_rex" "1")
(set_attr "mode" "")
(set_attr "athlon_decode" "double,vector")
(set_attr "amdfam10_decode" "double,double")])
(define_insn "fix_truncsi_sse"
[(set (match_operand:SI 0 "register_operand" "=r,r")
(fix:SI (match_operand:MODEF 1 "nonimmediate_operand" "x,m")))]
"SSE_FLOAT_MODE_P (mode)
&& (!TARGET_FISTTP || TARGET_SSE_MATH)"
"%vcvtts2si\t{%1, %0|%0, %1}"
[(set_attr "type" "sseicvt")
(set_attr "prefix" "maybe_vex")
(set_attr "mode" "")
(set_attr "athlon_decode" "double,vector")
(set_attr "amdfam10_decode" "double,double")])
;; Shorten x87->SSE reload sequences of fix_trunc?f?i_sse patterns.
(define_peephole2
[(set (match_operand:MODEF 0 "register_operand" "")
(match_operand:MODEF 1 "memory_operand" ""))
(set (match_operand:SSEMODEI24 2 "register_operand" "")
(fix:SSEMODEI24 (match_dup 0)))]
"TARGET_SHORTEN_X87_SSE
&& peep2_reg_dead_p (2, operands[0])"
[(set (match_dup 2) (fix:SSEMODEI24 (match_dup 1)))]
"")
;; Avoid vector decoded forms of the instruction.
(define_peephole2
[(match_scratch:DF 2 "Y2")
(set (match_operand:SSEMODEI24 0 "register_operand" "")
(fix:SSEMODEI24 (match_operand:DF 1 "memory_operand" "")))]
"TARGET_AVOID_VECTOR_DECODE && optimize_insn_for_speed_p ()"
[(set (match_dup 2) (match_dup 1))
(set (match_dup 0) (fix:SSEMODEI24 (match_dup 2)))]
"")
(define_peephole2
[(match_scratch:SF 2 "x")
(set (match_operand:SSEMODEI24 0 "register_operand" "")
(fix:SSEMODEI24 (match_operand:SF 1 "memory_operand" "")))]
"TARGET_AVOID_VECTOR_DECODE && optimize_insn_for_speed_p ()"
[(set (match_dup 2) (match_dup 1))
(set (match_dup 0) (fix:SSEMODEI24 (match_dup 2)))]
"")
(define_insn_and_split "fix_trunc_fisttp_i387_1"
[(set (match_operand:X87MODEI 0 "nonimmediate_operand" "")
(fix:X87MODEI (match_operand 1 "register_operand" "")))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& TARGET_FISTTP
&& !((SSE_FLOAT_MODE_P (GET_MODE (operands[1]))
&& (TARGET_64BIT || mode != DImode))
&& TARGET_SSE_MATH)
&& can_create_pseudo_p ()"
"#"
"&& 1"
[(const_int 0)]
{
if (memory_operand (operands[0], VOIDmode))
emit_insn (gen_fix_trunc_i387_fisttp (operands[0], operands[1]));
else
{
operands[2] = assign_386_stack_local (mode, SLOT_TEMP);
emit_insn (gen_fix_trunc_i387_fisttp_with_temp (operands[0],
operands[1],
operands[2]));
}
DONE;
}
[(set_attr "type" "fisttp")
(set_attr "mode" "")])
(define_insn "fix_trunc_i387_fisttp"
[(set (match_operand:X87MODEI 0 "memory_operand" "=m")
(fix:X87MODEI (match_operand 1 "register_operand" "f")))
(clobber (match_scratch:XF 2 "=&1f"))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& TARGET_FISTTP
&& !((SSE_FLOAT_MODE_P (GET_MODE (operands[1]))
&& (TARGET_64BIT || mode != DImode))
&& TARGET_SSE_MATH)"
"* return output_fix_trunc (insn, operands, 1);"
[(set_attr "type" "fisttp")
(set_attr "mode" "")])
(define_insn "fix_trunc_i387_fisttp_with_temp"
[(set (match_operand:X87MODEI 0 "nonimmediate_operand" "=m,?r")
(fix:X87MODEI (match_operand 1 "register_operand" "f,f")))
(clobber (match_operand:X87MODEI 2 "memory_operand" "=X,m"))
(clobber (match_scratch:XF 3 "=&1f,&1f"))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& TARGET_FISTTP
&& !((SSE_FLOAT_MODE_P (GET_MODE (operands[1]))
&& (TARGET_64BIT || mode != DImode))
&& TARGET_SSE_MATH)"
"#"
[(set_attr "type" "fisttp")
(set_attr "mode" "")])
(define_split
[(set (match_operand:X87MODEI 0 "register_operand" "")
(fix:X87MODEI (match_operand 1 "register_operand" "")))
(clobber (match_operand:X87MODEI 2 "memory_operand" ""))
(clobber (match_scratch 3 ""))]
"reload_completed"
[(parallel [(set (match_dup 2) (fix:X87MODEI (match_dup 1)))
(clobber (match_dup 3))])
(set (match_dup 0) (match_dup 2))]
"")
(define_split
[(set (match_operand:X87MODEI 0 "memory_operand" "")
(fix:X87MODEI (match_operand 1 "register_operand" "")))
(clobber (match_operand:X87MODEI 2 "memory_operand" ""))
(clobber (match_scratch 3 ""))]
"reload_completed"
[(parallel [(set (match_dup 0) (fix:X87MODEI (match_dup 1)))
(clobber (match_dup 3))])]
"")
;; See the comments in i386.h near OPTIMIZE_MODE_SWITCHING for the description
;; of the machinery. Please note the clobber of FLAGS_REG. In i387 control
;; word calculation (inserted by LCM in mode switching pass) a FLAGS_REG
;; clobbering insns can be used. Look at emit_i387_cw_initialization ()
;; function in i386.c.
(define_insn_and_split "*fix_trunc_i387_1"
[(set (match_operand:X87MODEI 0 "nonimmediate_operand" "")
(fix:X87MODEI (match_operand 1 "register_operand" "")))
(clobber (reg:CC FLAGS_REG))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& !TARGET_FISTTP
&& !(SSE_FLOAT_MODE_P (GET_MODE (operands[1]))
&& (TARGET_64BIT || mode != DImode))
&& can_create_pseudo_p ()"
"#"
"&& 1"
[(const_int 0)]
{
ix86_optimize_mode_switching[I387_TRUNC] = 1;
operands[2] = assign_386_stack_local (HImode, SLOT_CW_STORED);
operands[3] = assign_386_stack_local (HImode, SLOT_CW_TRUNC);
if (memory_operand (operands[0], VOIDmode))
emit_insn (gen_fix_trunc_i387 (operands[0], operands[1],
operands[2], operands[3]));
else
{
operands[4] = assign_386_stack_local (mode, SLOT_TEMP);
emit_insn (gen_fix_trunc_i387_with_temp (operands[0], operands[1],
operands[2], operands[3],
operands[4]));
}
DONE;
}
[(set_attr "type" "fistp")
(set_attr "i387_cw" "trunc")
(set_attr "mode" "")])
(define_insn "fix_truncdi_i387"
[(set (match_operand:DI 0 "memory_operand" "=m")
(fix:DI (match_operand 1 "register_operand" "f")))
(use (match_operand:HI 2 "memory_operand" "m"))
(use (match_operand:HI 3 "memory_operand" "m"))
(clobber (match_scratch:XF 4 "=&1f"))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& !TARGET_FISTTP
&& !(TARGET_64BIT && SSE_FLOAT_MODE_P (GET_MODE (operands[1])))"
"* return output_fix_trunc (insn, operands, 0);"
[(set_attr "type" "fistp")
(set_attr "i387_cw" "trunc")
(set_attr "mode" "DI")])
(define_insn "fix_truncdi_i387_with_temp"
[(set (match_operand:DI 0 "nonimmediate_operand" "=m,?r")
(fix:DI (match_operand 1 "register_operand" "f,f")))
(use (match_operand:HI 2 "memory_operand" "m,m"))
(use (match_operand:HI 3 "memory_operand" "m,m"))
(clobber (match_operand:DI 4 "memory_operand" "=X,m"))
(clobber (match_scratch:XF 5 "=&1f,&1f"))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& !TARGET_FISTTP
&& !(TARGET_64BIT && SSE_FLOAT_MODE_P (GET_MODE (operands[1])))"
"#"
[(set_attr "type" "fistp")
(set_attr "i387_cw" "trunc")
(set_attr "mode" "DI")])
(define_split
[(set (match_operand:DI 0 "register_operand" "")
(fix:DI (match_operand 1 "register_operand" "")))
(use (match_operand:HI 2 "memory_operand" ""))
(use (match_operand:HI 3 "memory_operand" ""))
(clobber (match_operand:DI 4 "memory_operand" ""))
(clobber (match_scratch 5 ""))]
"reload_completed"
[(parallel [(set (match_dup 4) (fix:DI (match_dup 1)))
(use (match_dup 2))
(use (match_dup 3))
(clobber (match_dup 5))])
(set (match_dup 0) (match_dup 4))]
"")
(define_split
[(set (match_operand:DI 0 "memory_operand" "")
(fix:DI (match_operand 1 "register_operand" "")))
(use (match_operand:HI 2 "memory_operand" ""))
(use (match_operand:HI 3 "memory_operand" ""))
(clobber (match_operand:DI 4 "memory_operand" ""))
(clobber (match_scratch 5 ""))]
"reload_completed"
[(parallel [(set (match_dup 0) (fix:DI (match_dup 1)))
(use (match_dup 2))
(use (match_dup 3))
(clobber (match_dup 5))])]
"")
(define_insn "fix_trunc_i387"
[(set (match_operand:X87MODEI12 0 "memory_operand" "=m")
(fix:X87MODEI12 (match_operand 1 "register_operand" "f")))
(use (match_operand:HI 2 "memory_operand" "m"))
(use (match_operand:HI 3 "memory_operand" "m"))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& !TARGET_FISTTP
&& !SSE_FLOAT_MODE_P (GET_MODE (operands[1]))"
"* return output_fix_trunc (insn, operands, 0);"
[(set_attr "type" "fistp")
(set_attr "i387_cw" "trunc")
(set_attr "mode" "")])
(define_insn "fix_trunc_i387_with_temp"
[(set (match_operand:X87MODEI12 0 "nonimmediate_operand" "=m,?r")
(fix:X87MODEI12 (match_operand 1 "register_operand" "f,f")))
(use (match_operand:HI 2 "memory_operand" "m,m"))
(use (match_operand:HI 3 "memory_operand" "m,m"))
(clobber (match_operand:X87MODEI12 4 "memory_operand" "=X,m"))]
"X87_FLOAT_MODE_P (GET_MODE (operands[1]))
&& !TARGET_FISTTP
&& !SSE_FLOAT_MODE_P (GET_MODE (operands[1]))"
"#"
[(set_attr "type" "fistp")
(set_attr "i387_cw" "trunc")
(set_attr "mode" "")])
(define_split
[(set (match_operand:X87MODEI12 0 "register_operand" "")
(fix:X87MODEI12 (match_operand 1 "register_operand" "")))
(use (match_operand:HI 2 "memory_operand" ""))
(use (match_operand:HI 3 "memory_operand" ""))
(clobber (match_operand:X87MODEI12 4 "memory_operand" ""))]
"reload_completed"
[(parallel [(set (match_dup 4) (fix:X87MODEI12 (match_dup 1)))
(use (match_dup 2))
(use (match_dup 3))])
(set (match_dup 0) (match_dup 4))]
"")
(define_split
[(set (match_operand:X87MODEI12 0 "memory_operand" "")
(fix:X87MODEI12 (match_operand 1 "register_operand" "")))
(use (match_operand:HI 2 "memory_operand" ""))
(use (match_operand:HI 3 "memory_operand" ""))
(clobber (match_operand:X87MODEI12 4 "memory_operand" ""))]
"reload_completed"
[(parallel [(set (match_dup 0) (fix:X87MODEI12 (match_dup 1)))
(use (match_dup 2))
(use (match_dup 3))])]
"")
(define_insn "x86_fnstcw_1"
[(set (match_operand:HI 0 "memory_operand" "=m")
(unspec:HI [(reg:HI FPCR_REG)] UNSPEC_FSTCW))]
"TARGET_80387"
"fnstcw\t%0"
[(set (attr "length") (symbol_ref "ix86_attr_length_address_default (insn) + 2"))
(set_attr "mode" "HI")
(set_attr "unit" "i387")])
(define_insn "x86_fldcw_1"
[(set (reg:HI FPCR_REG)
(unspec:HI [(match_operand:HI 0 "memory_operand" "m")] UNSPEC_FLDCW))]
"TARGET_80387"
"fldcw\t%0"
[(set (attr "length") (symbol_ref "ix86_attr_length_address_default (insn) + 2"))
(set_attr "mode" "HI")
(set_attr "unit" "i387")
(set_attr "athlon_decode" "vector")
(set_attr "amdfam10_decode" "vector")])
;; Conversion between fixed point and floating point.
;; Even though we only accept memory inputs, the backend _really_
;; wants to be able to do this between registers.
(define_expand "floathi2"
[(set (match_operand:X87MODEF 0 "register_operand" "")
(float:X87MODEF (match_operand:HI 1 "nonimmediate_operand" "")))]
"TARGET_80387
&& (!(SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH)
|| TARGET_MIX_SSE_I387)"
"")
;; Pre-reload splitter to add memory clobber to the pattern.
(define_insn_and_split "*floathi2_1"
[(set (match_operand:X87MODEF 0 "register_operand" "")
(float:X87MODEF (match_operand:HI 1 "register_operand" "")))]
"TARGET_80387
&& (!(SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH)
|| TARGET_MIX_SSE_I387)
&& can_create_pseudo_p ()"
"#"
"&& 1"
[(parallel [(set (match_dup 0)
(float:X87MODEF (match_dup 1)))
(clobber (match_dup 2))])]
"operands[2] = assign_386_stack_local (HImode, SLOT_TEMP);")
(define_insn "*floathi2_i387_with_temp"
[(set (match_operand:X87MODEF 0 "register_operand" "=f,f")
(float:X87MODEF (match_operand:HI 1 "nonimmediate_operand" "m,?r")))
(clobber (match_operand:HI 2 "memory_operand" "=m,m"))]
"TARGET_80387
&& (!(SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH)
|| TARGET_MIX_SSE_I387)"
"#"
[(set_attr "type" "fmov,multi")
(set_attr "mode" "")
(set_attr "unit" "*,i387")
(set_attr "fp_int_src" "true")])
(define_insn "*floathi2_i387"
[(set (match_operand:X87MODEF 0 "register_operand" "=f")
(float:X87MODEF (match_operand:HI 1 "memory_operand" "m")))]
"TARGET_80387
&& (!(SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH)
|| TARGET_MIX_SSE_I387)"
"fild%Z1\t%1"
[(set_attr "type" "fmov")
(set_attr "mode" "")
(set_attr "fp_int_src" "true")])
(define_split
[(set (match_operand:X87MODEF 0 "register_operand" "")
(float:X87MODEF (match_operand:HI 1 "register_operand" "")))
(clobber (match_operand:HI 2 "memory_operand" ""))]
"TARGET_80387
&& (!(SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH)
|| TARGET_MIX_SSE_I387)
&& reload_completed"
[(set (match_dup 2) (match_dup 1))
(set (match_dup 0) (float:X87MODEF (match_dup 2)))]
"")
(define_split
[(set (match_operand:X87MODEF 0 "register_operand" "")
(float:X87MODEF (match_operand:HI 1 "memory_operand" "")))
(clobber (match_operand:HI 2 "memory_operand" ""))]
"TARGET_80387
&& (!(SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH)
|| TARGET_MIX_SSE_I387)
&& reload_completed"
[(set (match_dup 0) (float:X87MODEF (match_dup 1)))]
"")
(define_expand "float2"
[(set (match_operand:X87MODEF 0 "register_operand" "")
(float:X87MODEF
(match_operand:SSEMODEI24 1 "nonimmediate_operand" "")))]
"TARGET_80387
|| ((mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH)"
"
{
if (!((mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH)
&& !X87_ENABLE_FLOAT (mode, mode))
{
rtx reg = gen_reg_rtx (XFmode);
rtx insn;
emit_insn (gen_floatxf2 (reg, operands[1]));
if (mode == SFmode)
insn = gen_truncxfsf2 (operands[0], reg);
else if (mode == DFmode)
insn = gen_truncxfdf2 (operands[0], reg);
else
gcc_unreachable ();
emit_insn (insn);
DONE;
}
}")
;; Pre-reload splitter to add memory clobber to the pattern.
(define_insn_and_split "*float2_1"
[(set (match_operand:X87MODEF 0 "register_operand" "")
(float:X87MODEF (match_operand:SSEMODEI24 1 "register_operand" "")))]
"((TARGET_80387
&& X87_ENABLE_FLOAT (mode, mode)
&& (!((mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH)
|| TARGET_MIX_SSE_I387))
|| ((mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH
&& ((mode == SImode
&& TARGET_SSE2 && TARGET_USE_VECTOR_CONVERTS
&& optimize_function_for_speed_p (cfun)
&& flag_trapping_math)
|| !(TARGET_INTER_UNIT_CONVERSIONS
|| optimize_function_for_size_p (cfun)))))
&& can_create_pseudo_p ()"
"#"
"&& 1"
[(parallel [(set (match_dup 0) (float:X87MODEF (match_dup 1)))
(clobber (match_dup 2))])]
{
operands[2] = assign_386_stack_local (mode, SLOT_TEMP);
/* Avoid store forwarding (partial memory) stall penalty
by passing DImode value through XMM registers. */
if (mode == DImode && !TARGET_64BIT
&& TARGET_80387 && TARGET_SSE2 && TARGET_INTER_UNIT_MOVES
&& optimize_function_for_speed_p (cfun))
{
emit_insn (gen_floatdi2_i387_with_xmm (operands[0],
operands[1],
operands[2]));
DONE;
}
})
(define_insn "*floatsi2_vector_mixed_with_temp"
[(set (match_operand:MODEF 0 "register_operand" "=f,f,x,x,x")
(float:MODEF
(match_operand:SI 1 "nonimmediate_operand" "m,?r,r,m,!x")))
(clobber (match_operand:SI 2 "memory_operand" "=X,m,m,X,m"))]
"TARGET_SSE2 && TARGET_MIX_SSE_I387
&& TARGET_USE_VECTOR_CONVERTS && optimize_function_for_speed_p (cfun)"
"#"
[(set_attr "type" "fmov,multi,sseicvt,sseicvt,sseicvt")
(set_attr "mode" ",,,,")
(set_attr "unit" "*,i387,*,*,*")
(set_attr "athlon_decode" "*,*,double,direct,double")
(set_attr "amdfam10_decode" "*,*,vector,double,double")
(set_attr "fp_int_src" "true")])
(define_insn "*floatsi2_vector_mixed"
[(set (match_operand:MODEF 0 "register_operand" "=f,x")
(float:MODEF (match_operand:SI 1 "memory_operand" "m,m")))]
"TARGET_SSE2 && TARGET_MIX_SSE_I387
&& TARGET_USE_VECTOR_CONVERTS && optimize_function_for_speed_p (cfun)"
"@
fild%Z1\t%1
#"
[(set_attr "type" "fmov,sseicvt")
(set_attr "mode" ",")
(set_attr "unit" "i387,*")
(set_attr "athlon_decode" "*,direct")
(set_attr "amdfam10_decode" "*,double")
(set_attr "fp_int_src" "true")])
(define_insn "*float2_mixed_with_temp"
[(set (match_operand:MODEF 0 "register_operand" "=f,f,x,x")
(float:MODEF
(match_operand:SSEMODEI24 1 "nonimmediate_operand" "m,?r,r,m")))
(clobber (match_operand:SSEMODEI24 2 "memory_operand" "=X,m,m,X"))]
"(mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_MIX_SSE_I387"
"#"
[(set_attr "type" "fmov,multi,sseicvt,sseicvt")
(set_attr "mode" "")
(set_attr "unit" "*,i387,*,*")
(set_attr "athlon_decode" "*,*,double,direct")
(set_attr "amdfam10_decode" "*,*,vector,double")
(set_attr "fp_int_src" "true")])
(define_split
[(set (match_operand:MODEF 0 "register_operand" "")
(float:MODEF (match_operand:SSEMODEI24 1 "register_operand" "")))
(clobber (match_operand:SSEMODEI24 2 "memory_operand" ""))]
"(mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_MIX_SSE_I387
&& TARGET_INTER_UNIT_CONVERSIONS
&& reload_completed
&& (SSE_REG_P (operands[0])
|| (GET_CODE (operands[0]) == SUBREG
&& SSE_REG_P (operands[0])))"
[(set (match_dup 0) (float:MODEF (match_dup 1)))]
"")
(define_split
[(set (match_operand:MODEF 0 "register_operand" "")
(float:MODEF (match_operand:SSEMODEI24 1 "register_operand" "")))
(clobber (match_operand:SSEMODEI24 2 "memory_operand" ""))]
"(mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_MIX_SSE_I387
&& !(TARGET_INTER_UNIT_CONVERSIONS || optimize_function_for_size_p (cfun))
&& reload_completed
&& (SSE_REG_P (operands[0])
|| (GET_CODE (operands[0]) == SUBREG
&& SSE_REG_P (operands[0])))"
[(set (match_dup 2) (match_dup 1))
(set (match_dup 0) (float:MODEF (match_dup 2)))]
"")
(define_insn "*float2_mixed_interunit"
[(set (match_operand:MODEF 0 "register_operand" "=f,x,x")
(float:MODEF
(match_operand:SSEMODEI24 1 "nonimmediate_operand" "m,r,m")))]
"(mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_MIX_SSE_I387
&& (TARGET_INTER_UNIT_CONVERSIONS || optimize_function_for_size_p (cfun))"
"@
fild%Z1\t%1
%vcvtsi2s\t{%1, %d0|%d0, %1}
%vcvtsi2s\t{%1, %d0|%d0, %1}"
[(set_attr "type" "fmov,sseicvt,sseicvt")
(set_attr "prefix" "orig,maybe_vex,maybe_vex")
(set_attr "mode" "")
(set (attr "prefix_rex")
(if_then_else
(and (eq_attr "prefix" "maybe_vex")
(ne (symbol_ref "mode == DImode") (const_int 0)))
(const_string "1")
(const_string "*")))
(set_attr "unit" "i387,*,*")
(set_attr "athlon_decode" "*,double,direct")
(set_attr "amdfam10_decode" "*,vector,double")
(set_attr "fp_int_src" "true")])
(define_insn "*float2_mixed_nointerunit"
[(set (match_operand:MODEF 0 "register_operand" "=f,x")
(float:MODEF
(match_operand:SSEMODEI24 1 "memory_operand" "m,m")))]
"(mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_MIX_SSE_I387
&& !(TARGET_INTER_UNIT_CONVERSIONS || optimize_function_for_size_p (cfun))"
"@
fild%Z1\t%1
%vcvtsi2s\t{%1, %d0|%d0, %1}"
[(set_attr "type" "fmov,sseicvt")
(set_attr "prefix" "orig,maybe_vex")
(set_attr "mode" "")
(set (attr "prefix_rex")
(if_then_else
(and (eq_attr "prefix" "maybe_vex")
(ne (symbol_ref "mode == DImode") (const_int 0)))
(const_string "1")
(const_string "*")))
(set_attr "athlon_decode" "*,direct")
(set_attr "amdfam10_decode" "*,double")
(set_attr "fp_int_src" "true")])
(define_insn "*floatsi2_vector_sse_with_temp"
[(set (match_operand:MODEF 0 "register_operand" "=x,x,x")
(float:MODEF
(match_operand:SI 1 "nonimmediate_operand" "r,m,!x")))
(clobber (match_operand:SI 2 "memory_operand" "=m,X,m"))]
"TARGET_SSE2 && TARGET_SSE_MATH
&& TARGET_USE_VECTOR_CONVERTS && optimize_function_for_speed_p (cfun)"
"#"
[(set_attr "type" "sseicvt")
(set_attr "mode" ",,")
(set_attr "athlon_decode" "double,direct,double")
(set_attr "amdfam10_decode" "vector,double,double")
(set_attr "fp_int_src" "true")])
(define_insn "*floatsi2_vector_sse"
[(set (match_operand:MODEF 0 "register_operand" "=x")
(float:MODEF (match_operand:SI 1 "memory_operand" "m")))]
"TARGET_SSE2 && TARGET_SSE_MATH
&& TARGET_USE_VECTOR_CONVERTS && optimize_function_for_speed_p (cfun)"
"#"
[(set_attr "type" "sseicvt")
(set_attr "mode" "")
(set_attr "athlon_decode" "direct")
(set_attr "amdfam10_decode" "double")
(set_attr "fp_int_src" "true")])
(define_split
[(set (match_operand:MODEF 0 "register_operand" "")
(float:MODEF (match_operand:SI 1 "register_operand" "")))
(clobber (match_operand:SI 2 "memory_operand" ""))]
"TARGET_SSE2 && TARGET_SSE_MATH
&& TARGET_USE_VECTOR_CONVERTS && optimize_function_for_speed_p (cfun)
&& reload_completed
&& (SSE_REG_P (operands[0])
|| (GET_CODE (operands[0]) == SUBREG
&& SSE_REG_P (operands[0])))"
[(const_int 0)]
{
rtx op1 = operands[1];
operands[3] = simplify_gen_subreg (mode, operands[0],
mode, 0);
if (GET_CODE (op1) == SUBREG)
op1 = SUBREG_REG (op1);
if (GENERAL_REG_P (op1) && TARGET_INTER_UNIT_MOVES)
{
operands[4] = simplify_gen_subreg (V4SImode, operands[0], mode, 0);
emit_insn (gen_sse2_loadld (operands[4],
CONST0_RTX (V4SImode), operands[1]));
}
/* We can ignore possible trapping value in the
high part of SSE register for non-trapping math. */
else if (SSE_REG_P (op1) && !flag_trapping_math)
operands[4] = simplify_gen_subreg (V4SImode, operands[1], SImode, 0);
else
{
operands[4] = simplify_gen_subreg (V4SImode, operands[0], mode, 0);
emit_move_insn (operands[2], operands[1]);
emit_insn (gen_sse2_loadld (operands[4],
CONST0_RTX (V4SImode), operands[2]));
}
emit_insn
(gen_sse2_cvtdq2p (operands[3], operands[4]));
DONE;
})
(define_split
[(set (match_operand:MODEF 0 "register_operand" "")
(float:MODEF (match_operand:SI 1 "memory_operand" "")))
(clobber (match_operand:SI 2 "memory_operand" ""))]
"TARGET_SSE2 && TARGET_SSE_MATH
&& TARGET_USE_VECTOR_CONVERTS && optimize_function_for_speed_p (cfun)
&& reload_completed
&& (SSE_REG_P (operands[0])
|| (GET_CODE (operands[0]) == SUBREG
&& SSE_REG_P (operands[0])))"
[(const_int 0)]
{
operands[3] = simplify_gen_subreg (mode, operands[0],
mode, 0);
operands[4] = simplify_gen_subreg (V4SImode, operands[0], mode, 0);
emit_insn (gen_sse2_loadld (operands[4],
CONST0_RTX (V4SImode), operands[1]));
emit_insn
(gen_sse2_cvtdq2p (operands[3], operands[4]));
DONE;
})
(define_split
[(set (match_operand:MODEF 0 "register_operand" "")
(float:MODEF (match_operand:SI 1 "register_operand" "")))]
"TARGET_SSE2 && TARGET_SSE_MATH
&& TARGET_USE_VECTOR_CONVERTS && optimize_function_for_speed_p (cfun)
&& reload_completed
&& (SSE_REG_P (operands[0])
|| (GET_CODE (operands[0]) == SUBREG
&& SSE_REG_P (operands[0])))"
[(const_int 0)]
{
rtx op1 = operands[1];
operands[3] = simplify_gen_subreg (mode, operands[0],
mode, 0);
if (GET_CODE (op1) == SUBREG)
op1 = SUBREG_REG (op1);
if (GENERAL_REG_P (op1) && TARGET_INTER_UNIT_MOVES)
{
operands[4] = simplify_gen_subreg (V4SImode, operands[0], mode, 0);
emit_insn (gen_sse2_loadld (operands[4],
CONST0_RTX (V4SImode), operands[1]));
}
/* We can ignore possible trapping value in the
high part of SSE register for non-trapping math. */
else if (SSE_REG_P (op1) && !flag_trapping_math)
operands[4] = simplify_gen_subreg (V4SImode, operands[1], SImode, 0);
else
gcc_unreachable ();
emit_insn
(gen_sse2_cvtdq2p (operands[3], operands[4]));
DONE;
})
(define_split
[(set (match_operand:MODEF 0 "register_operand" "")
(float:MODEF (match_operand:SI 1 "memory_operand" "")))]
"TARGET_SSE2 && TARGET_SSE_MATH
&& TARGET_USE_VECTOR_CONVERTS && optimize_function_for_speed_p (cfun)
&& reload_completed
&& (SSE_REG_P (operands[0])
|| (GET_CODE (operands[0]) == SUBREG
&& SSE_REG_P (operands[0])))"
[(const_int 0)]
{
operands[3] = simplify_gen_subreg (mode, operands[0],
mode, 0);
operands[4] = simplify_gen_subreg (V4SImode, operands[0], mode, 0);
emit_insn (gen_sse2_loadld (operands[4],
CONST0_RTX (V4SImode), operands[1]));
emit_insn
(gen_sse2_cvtdq2p (operands[3], operands[4]));
DONE;
})
(define_insn "*float2_sse_with_temp"
[(set (match_operand:MODEF 0 "register_operand" "=x,x")
(float:MODEF
(match_operand:SSEMODEI24 1 "nonimmediate_operand" "r,m")))
(clobber (match_operand:SSEMODEI24 2 "memory_operand" "=m,X"))]
"(mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH"
"#"
[(set_attr "type" "sseicvt")
(set_attr "mode" "")
(set_attr "athlon_decode" "double,direct")
(set_attr "amdfam10_decode" "vector,double")
(set_attr "fp_int_src" "true")])
(define_insn "*float2_sse_interunit"
[(set (match_operand:MODEF 0 "register_operand" "=x,x")
(float:MODEF
(match_operand:SSEMODEI24 1 "nonimmediate_operand" "r,m")))]
"(mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH
&& (TARGET_INTER_UNIT_CONVERSIONS || optimize_function_for_size_p (cfun))"
"%vcvtsi2s\t{%1, %d0|%d0, %1}"
[(set_attr "type" "sseicvt")
(set_attr "prefix" "maybe_vex")
(set_attr "mode" "")
(set (attr "prefix_rex")
(if_then_else
(and (eq_attr "prefix" "maybe_vex")
(ne (symbol_ref "mode == DImode") (const_int 0)))
(const_string "1")
(const_string "*")))
(set_attr "athlon_decode" "double,direct")
(set_attr "amdfam10_decode" "vector,double")
(set_attr "fp_int_src" "true")])
(define_split
[(set (match_operand:MODEF 0 "register_operand" "")
(float:MODEF (match_operand:SSEMODEI24 1 "nonimmediate_operand" "")))
(clobber (match_operand:SSEMODEI24 2 "memory_operand" ""))]
"(mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH
&& (TARGET_INTER_UNIT_CONVERSIONS || optimize_function_for_size_p (cfun))
&& reload_completed
&& (SSE_REG_P (operands[0])
|| (GET_CODE (operands[0]) == SUBREG
&& SSE_REG_P (operands[0])))"
[(set (match_dup 0) (float:MODEF (match_dup 1)))]
"")
(define_insn "*float2_sse_nointerunit"
[(set (match_operand:MODEF 0 "register_operand" "=x")
(float:MODEF
(match_operand:SSEMODEI24 1 "memory_operand" "m")))]
"(mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH
&& !(TARGET_INTER_UNIT_CONVERSIONS || optimize_function_for_size_p (cfun))"
"%vcvtsi2s\t{%1, %d0|%d0, %1}"
[(set_attr "type" "sseicvt")
(set_attr "prefix" "maybe_vex")
(set_attr "mode" "")
(set (attr "prefix_rex")
(if_then_else
(and (eq_attr "prefix" "maybe_vex")
(ne (symbol_ref "mode == DImode") (const_int 0)))
(const_string "1")
(const_string "*")))
(set_attr "athlon_decode" "direct")
(set_attr "amdfam10_decode" "double")
(set_attr "fp_int_src" "true")])
(define_split
[(set (match_operand:MODEF 0 "register_operand" "")
(float:MODEF (match_operand:SSEMODEI24 1 "register_operand" "")))
(clobber (match_operand:SSEMODEI24 2 "memory_operand" ""))]
"(mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH
&& !(TARGET_INTER_UNIT_CONVERSIONS || optimize_function_for_size_p (cfun))
&& reload_completed
&& (SSE_REG_P (operands[0])
|| (GET_CODE (operands[0]) == SUBREG
&& SSE_REG_P (operands[0])))"
[(set (match_dup 2) (match_dup 1))
(set (match_dup 0) (float:MODEF (match_dup 2)))]
"")
(define_split
[(set (match_operand:MODEF 0 "register_operand" "")
(float:MODEF (match_operand:SSEMODEI24 1 "memory_operand" "")))
(clobber (match_operand:SSEMODEI24 2 "memory_operand" ""))]
"(mode != DImode || TARGET_64BIT)
&& SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH
&& reload_completed
&& (SSE_REG_P (operands[0])
|| (GET_CODE (operands[0]) == SUBREG
&& SSE_REG_P (operands[0])))"
[(set (match_dup 0) (float:MODEF (match_dup 1)))]
"")
(define_insn "*float2_i387_with_temp"
[(set (match_operand:X87MODEF 0 "register_operand" "=f,f")
(float:X87MODEF
(match_operand:SSEMODEI24 1 "nonimmediate_operand" "m,?r")))
(clobber (match_operand:SSEMODEI24 2 "memory_operand" "=X,m"))]
"TARGET_80387
&& X87_ENABLE_FLOAT (mode, mode)"
"@
fild%Z1\t%1
#"
[(set_attr "type" "fmov,multi")
(set_attr "mode" "")
(set_attr "unit" "*,i387")
(set_attr "fp_int_src" "true")])
(define_insn "*float2_i387"
[(set (match_operand:X87MODEF 0 "register_operand" "=f")
(float:X87MODEF
(match_operand:SSEMODEI24 1 "memory_operand" "m")))]
"TARGET_80387
&& X87_ENABLE_FLOAT (mode, mode)"
"fild%Z1\t%1"
[(set_attr "type" "fmov")
(set_attr "mode" "")
(set_attr "fp_int_src" "true")])
(define_split
[(set (match_operand:X87MODEF 0 "register_operand" "")
(float:X87MODEF (match_operand:SSEMODEI24 1 "register_operand" "")))
(clobber (match_operand:SSEMODEI24 2 "memory_operand" ""))]
"TARGET_80387
&& X87_ENABLE_FLOAT (mode, mode)
&& reload_completed
&& FP_REG_P (operands[0])"
[(set (match_dup 2) (match_dup 1))
(set (match_dup 0) (float:X87MODEF (match_dup 2)))]
"")
(define_split
[(set (match_operand:X87MODEF 0 "register_operand" "")
(float:X87MODEF (match_operand:SSEMODEI24 1 "memory_operand" "")))
(clobber (match_operand:SSEMODEI24 2 "memory_operand" ""))]
"TARGET_80387
&& X87_ENABLE_FLOAT (mode, mode)
&& reload_completed
&& FP_REG_P (operands[0])"
[(set (match_dup 0) (float:X87MODEF (match_dup 1)))]
"")
;; Avoid store forwarding (partial memory) stall penalty
;; by passing DImode value through XMM registers. */
(define_insn "floatdi2_i387_with_xmm"
[(set (match_operand:X87MODEF 0 "register_operand" "=f,f")
(float:X87MODEF
(match_operand:DI 1 "nonimmediate_operand" "m,?r")))
(clobber (match_scratch:V4SI 3 "=X,x"))
(clobber (match_scratch:V4SI 4 "=X,x"))
(clobber (match_operand:DI 2 "memory_operand" "=X,m"))]
"TARGET_80387 && X87_ENABLE_FLOAT (mode, DImode)
&& TARGET_SSE2 && TARGET_INTER_UNIT_MOVES
&& !TARGET_64BIT && optimize_function_for_speed_p (cfun)"
"#"
[(set_attr "type" "multi")
(set_attr "mode" "")
(set_attr "unit" "i387")
(set_attr "fp_int_src" "true")])
(define_split
[(set (match_operand:X87MODEF 0 "register_operand" "")
(float:X87MODEF (match_operand:DI 1 "register_operand" "")))
(clobber (match_scratch:V4SI 3 ""))
(clobber (match_scratch:V4SI 4 ""))
(clobber (match_operand:DI 2 "memory_operand" ""))]
"TARGET_80387 && X87_ENABLE_FLOAT (mode, DImode)
&& TARGET_SSE2 && TARGET_INTER_UNIT_MOVES
&& !TARGET_64BIT && optimize_function_for_speed_p (cfun)
&& reload_completed
&& FP_REG_P (operands[0])"
[(set (match_dup 2) (match_dup 3))
(set (match_dup 0) (float:X87MODEF (match_dup 2)))]
{
/* The DImode arrived in a pair of integral registers (e.g. %edx:%eax).
Assemble the 64-bit DImode value in an xmm register. */
emit_insn (gen_sse2_loadld (operands[3], CONST0_RTX (V4SImode),
gen_rtx_SUBREG (SImode, operands[1], 0)));
emit_insn (gen_sse2_loadld (operands[4], CONST0_RTX (V4SImode),
gen_rtx_SUBREG (SImode, operands[1], 4)));
emit_insn (gen_sse2_punpckldq (operands[3], operands[3], operands[4]));
operands[3] = gen_rtx_REG (DImode, REGNO (operands[3]));
})
(define_split
[(set (match_operand:X87MODEF 0 "register_operand" "")
(float:X87MODEF (match_operand:DI 1 "memory_operand" "")))
(clobber (match_scratch:V4SI 3 ""))
(clobber (match_scratch:V4SI 4 ""))
(clobber (match_operand:DI 2 "memory_operand" ""))]
"TARGET_80387 && X87_ENABLE_FLOAT (mode, DImode)
&& TARGET_SSE2 && TARGET_INTER_UNIT_MOVES
&& !TARGET_64BIT && optimize_function_for_speed_p (cfun)
&& reload_completed
&& FP_REG_P (operands[0])"
[(set (match_dup 0) (float:X87MODEF (match_dup 1)))]
"")
;; Avoid store forwarding (partial memory) stall penalty by extending
;; SImode value to DImode through XMM register instead of pushing two
;; SImode values to stack. Note that even !TARGET_INTER_UNIT_MOVES
;; targets benefit from this optimization. Also note that fild
;; loads from memory only.
(define_insn "*floatunssi2_1"
[(set (match_operand:X87MODEF 0 "register_operand" "=f,f")
(unsigned_float:X87MODEF
(match_operand:SI 1 "nonimmediate_operand" "x,m")))
(clobber (match_operand:DI 2 "memory_operand" "=m,m"))
(clobber (match_scratch:SI 3 "=X,x"))]
"!TARGET_64BIT
&& TARGET_80387 && X87_ENABLE_FLOAT (mode, DImode)
&& TARGET_SSE"
"#"
[(set_attr "type" "multi")
(set_attr "mode" "")])
(define_split
[(set (match_operand:X87MODEF 0 "register_operand" "")
(unsigned_float:X87MODEF
(match_operand:SI 1 "register_operand" "")))
(clobber (match_operand:DI 2 "memory_operand" ""))
(clobber (match_scratch:SI 3 ""))]
"!TARGET_64BIT
&& TARGET_80387 && X87_ENABLE_FLOAT (mode, DImode)
&& TARGET_SSE
&& reload_completed"
[(set (match_dup 2) (match_dup 1))
(set (match_dup 0)
(float:X87MODEF (match_dup 2)))]
"operands[1] = simplify_gen_subreg (DImode, operands[1], SImode, 0);")
(define_split
[(set (match_operand:X87MODEF 0 "register_operand" "")
(unsigned_float:X87MODEF
(match_operand:SI 1 "memory_operand" "")))
(clobber (match_operand:DI 2 "memory_operand" ""))
(clobber (match_scratch:SI 3 ""))]
"!TARGET_64BIT
&& TARGET_80387 && X87_ENABLE_FLOAT (mode, DImode)
&& TARGET_SSE
&& reload_completed"
[(set (match_dup 2) (match_dup 3))
(set (match_dup 0)
(float:X87MODEF (match_dup 2)))]
{
emit_move_insn (operands[3], operands[1]);
operands[3] = simplify_gen_subreg (DImode, operands[3], SImode, 0);
})
(define_expand "floatunssi2"
[(parallel
[(set (match_operand:X87MODEF 0 "register_operand" "")
(unsigned_float:X87MODEF
(match_operand:SI 1 "nonimmediate_operand" "")))
(clobber (match_dup 2))
(clobber (match_scratch:SI 3 ""))])]
"!TARGET_64BIT
&& ((TARGET_80387 && X87_ENABLE_FLOAT (mode, DImode)
&& TARGET_SSE)
|| (SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH))"
{
if (SSE_FLOAT_MODE_P (mode) && TARGET_SSE_MATH)
{
ix86_expand_convert_uns_si_sse (operands[0], operands[1]);
DONE;
}
else
{
enum ix86_stack_slot slot = (virtuals_instantiated
? SLOT_TEMP
: SLOT_VIRTUAL);
operands[2] = assign_386_stack_local (DImode, slot);
}
})
(define_expand "floatunsdisf2"
[(use (match_operand:SF 0 "register_operand" ""))
(use (match_operand:DI 1 "nonimmediate_operand" ""))]
"TARGET_64BIT && TARGET_SSE_MATH"
"x86_emit_floatuns (operands); DONE;")
(define_expand "floatunsdidf2"
[(use (match_operand:DF 0 "register_operand" ""))
(use (match_operand:DI 1 "nonimmediate_operand" ""))]
"(TARGET_64BIT || TARGET_KEEPS_VECTOR_ALIGNED_STACK)
&& TARGET_SSE2 && TARGET_SSE_MATH"
{
if (TARGET_64BIT)
x86_emit_floatuns (operands);
else
ix86_expand_convert_uns_didf_sse (operands[0], operands[1]);
DONE;
})
;; Add instructions
(define_expand "add3"
[(set (match_operand:SDWIM 0 "nonimmediate_operand" "")
(plus:SDWIM (match_operand:SDWIM 1 "nonimmediate_operand" "")
(match_operand:SDWIM 2 "" "")))]
""
"ix86_expand_binary_operator (PLUS, mode, operands); DONE;")
(define_insn_and_split "*add3_doubleword"
[(set (match_operand: 0 "nonimmediate_operand" "=r,o")
(plus:
(match_operand: 1 "nonimmediate_operand" "%0,0")
(match_operand: 2 "" "ro,r")))
(clobber (reg:CC FLAGS_REG))]
"ix86_binary_operator_ok (PLUS, mode, operands)"
"#"
"reload_completed"
[(parallel [(set (reg:CC FLAGS_REG)
(unspec:CC [(match_dup 1) (match_dup 2)]
UNSPEC_ADD_CARRY))
(set (match_dup 0)
(plus:DWIH (match_dup 1) (match_dup 2)))])
(parallel [(set (match_dup 3)
(plus:DWIH
(plus:DWIH
(ltu:DWIH (reg:CC FLAGS_REG) (const_int 0))
(match_dup 4))
(match_dup 5)))
(clobber (reg:CC FLAGS_REG))])]
"split_ (&operands[0], 3, &operands[0], &operands[3]);")
(define_insn "add3_carry"
[(set (match_operand:SWI 0 "nonimmediate_operand" "=m,")
(plus:SWI
(plus:SWI (match_operand:SWI 3 "ix86_carry_flag_operator" "")
(match_operand:SWI 1 "nonimmediate_operand" "%0,0"))
(match_operand:SWI 2 "" ",m")))
(clobber (reg:CC FLAGS_REG))]
"ix86_binary_operator_ok (PLUS, mode, operands)"
"adc{