ROUND4(X0, X5, X15, X10, X1, X6, X12, X11, X2, X7, X13, X8,  X3, X4, X14, X9)

	ADD $-1, R1
	BNE loop

	// decrement length
	ADD $-256, R4

	// rearrange vectors
	SHUFFLE(X0, X1, X2, X3, M0, M1, M2, M3)
	ADDV(J0, X0, X1, X2, X3)
	SHUFFLE(X4, X5, X6, X7, M0, M1, M2, M3)
	ADDV(KEY0, X4, X5, X6, X7)
	SHUFFLE(X8, X9, X10, X11, M0, M1, M2, M3)
	ADDV(KEY1, X8, X9, X10, X11)

	MOVQ	R9, 24(SP)
	CALL	runtime·sehhandler(SB)
	MOVL	32(SP), AX

	ADJSP	$-40

	POP_REGS_HOST_TO_ABI0()
	RET

TEXT runtime·callbackasm1(SB),NOSPLIT|NOFRAME,$0
	// Construct args vector for cgocallback().
	// By windows/amd64 calling convention first 4 args are in CX, DX, R8, R9
	// args from the 5th on are on the stack.
	// In any case, even if function has 0,1,2,3,4 args, there is reserved

Examples:

	TEXT ·Add(SB),NOSPLIT|NOFRAME,$0
	  PCALIGN $2048
	  MOVD $1, R0
	  MOVD $1, R1
	  RET

7. Move large constants to vector registers.

Go asm uses VMOVQ/VMOVD/VMOVS to move 128-bit, 64-bit and 32-bit constants into vector registers, respectively.
And for a 128-bit integer, it take two 64-bit operands, for the low and high parts separately.

Examples:

	VMOVS $0x11223344, V0

	HasSHA512 bool // K{I,L}MD-SHA-512 functions
	HasSHA3   bool // K{I,L}MD-SHA3-{224,256,384,512} and K{I,L}MD-SHAKE-{128,256} functions
	HasVX     bool // vector facility. Note: the runtime sets this when it processes auxv records.
	HasVXE    bool // vector-enhancements facility 1
	HasKDSA   bool // elliptic curve functions
	HasECDSA  bool // NIST curves
	HasEDDSA  bool // Edwards curves
	_         CacheLinePad
}

	// A URL like `javascript:checkThatFormNotEditedBeforeLeavingPage()`
	// from a trusted source should go in the page, but by default dynamic
	// `javascript:` URLs are filtered out since they are a frequently
	// exploited injection vector.
	//
	// Use of this type presents a security risk:
	// the encapsulated content should come from a trusted source,
	// as it will be included verbatim in the template output.
	URL string

// license that can be found in the LICENSE file.

// backtrack is a regular expression search with submatch
// tracking for small regular expressions and texts. It allocates
// a bit vector with (length of input) * (length of prog) bits,
// to make sure it never explores the same (character position, instruction)
// state multiple times. This limits the search to run in time linear in
// the length of the test.
//

// license that can be found in the LICENSE file.

//go:build !math_big_pure_go && (mips || mipsle)

#include "textflag.h"

// This file provides fast assembly versions for the elementary
// arithmetic operations on vectors implemented in arith.go.

TEXT ·addVV(SB),NOSPLIT,$0
	JMP	·addVV_g(SB)

TEXT ·subVV(SB),NOSPLIT,$0
	JMP	·subVV_g(SB)

TEXT ·addVW(SB),NOSPLIT,$0
	JMP	·addVW_g(SB)

TEXT ·subVW(SB),NOSPLIT,$0

	"unsafe"
	_ "unsafe" // for go:linkname
)

// OS-specific startup can set startupRand if the OS passes
// random data to the process at startup time.
// For example Linux passes 16 bytes in the auxv vector.
var startupRand []byte

// globalRand holds the global random state.
// It is only used at startup and for creating new m's.
// Otherwise the per-m random state should be used
// by calling goodrand.

// license that can be found in the LICENSE file.

//go:build !math_big_pure_go

#include "textflag.h"

// This file provides fast assembly versions for the elementary
// arithmetic operations on vectors implemented in arith.go.

// func addVV(z, x, y []Word) (c Word)
TEXT ·addVV(SB),NOSPLIT,$0
	MOVL z+0(FP), DI
	MOVL x+12(FP), SI
	MOVL y+24(FP), CX
	MOVL z_len+4(FP), BP
	MOVL $0, BX		// i = 0

}

// Add returns the vector p+q.
func (p Point) Add(q Point) Point {
	return Point{p.X + q.X, p.Y + q.Y}
}

// Sub returns the vector p-q.
func (p Point) Sub(q Point) Point {
	return Point{p.X - q.X, p.Y - q.Y}
}

// Mul returns the vector p*k.
func (p Point) Mul(k int) Point {
	return Point{p.X * k, p.Y * k}
}

// Div returns the vector p/k.
func (p Point) Div(k int) Point {