// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.

//go:build riscv64 && gc

#include "textflag.h"

TEXT ·RewindAndSetgid(SB),NOSPLIT|NOFRAME,$0-0
	// Rewind stack pointer so anything that happens on the stack
	// will clobber the test pattern created by the caller
	ADD	$(1024*8), X2

	// Ask signaller to setgid
	MOV	$1, X5
	FENCE
	MOVW	X5, ·Baton(SB)
	FENCE

	ADD	Rc2, Rc
	ADD	Rc3, Rd

	MOVM.IA [Ra,Rb,Rc,Rd], (Rt0)

	MOVW	p_data, Rdata
	MOVW	p_end, Rt0
	ADD	$64, Rdata
	CMP	Rt0, Rdata
	BLO	loop

	RET

// MD5 constants table

	// Round 1
	DATA	·table+0x00(SB)/4, $0xd76aa478
	DATA	·table+0x04(SB)/4, $0xe8c7b756
	DATA	·table+0x08(SB)/4, $0x242070db
	DATA	·table+0x0c(SB)/4, $0xc1bdceee
	DATA	·table+0x10(SB)/4, $0xf57c0faf

// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.

//go:build (amd64 || amd64p32) && gc

#include "textflag.h"

TEXT ·RewindAndSetgid(SB),NOSPLIT,$0-0
	// Rewind stack pointer so anything that happens on the stack
	// will clobber the test pattern created by the caller
	ADDQ	$(1024 * 8), SP

	// Ask signaller to setgid
	MOVL	$1, ·Baton(SB)

	// Wait for setgid completion

//   - rounds 60-79 are type 4 and do not load data (ROUND4 macro).
//
// Each round loads or shuffles the data, then computes a per-round
// function of b, c, d, and then mixes the result into and rotates the
// five registers a, b, c, d, e holding the intermediate results.
//
// The register rotation is implemented by rotating the arguments to
// the round macros instead of by explicit move instructions.

#define LOAD(index) \

#define numBytes	DX	// input, number of input bytes to be processed
#define sha256Constants	AX	// round contents from K256 table, indexed by round number x 32
#define msg		X0	// input data
#define state0		X1	// round intermediates and outputs
#define state1		X2
#define m0		X3	// m0, m1,... m4 -- round message temps
#define m1		X4
#define m2		X5
#define m3		X6
#define m4		X7

	VAQ    T2, CAR2, T2

	// Third round
	VPERM T1, T0, SEL1, RED2    // d1 d0 d1 d0
	VPERM ZER, RED2, SEL2, RED1 // 0  d1 d0  0
	VSQ   RED1, RED2, RED2      // Guaranteed not to underflow

	VSLDB $8, T1, T0, T0
	VSLDB $8, T2, T1, T1

	VACCQ  T0, RED1, CAR1
	VAQ    T0, RED1, T0
	VACCCQ T1, RED2, CAR1, CAR2
	VACQ   T1, RED2, CAR1, T1
	VAQ    T2, CAR2, T2

	// Last round

// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.

//go:build (mips || mipsle) && gc

#include "textflag.h"

TEXT ·RewindAndSetgid(SB),NOSPLIT|NOFRAME,$0-0
	// Rewind stack pointer so anything that happens on the stack
	// will clobber the test pattern created by the caller
	ADDU	$(1024*8), R29

	// Ask signaller to setgid
	MOVW	$1, R1
	SYNC
	MOVW	R1, ·Baton(SB)
	SYNC

// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.

//go:build (ppc64 || ppc64le) && gc

#include "textflag.h"

TEXT ·RewindAndSetgid(SB),NOSPLIT|NOFRAME,$0-0
	// Rewind stack pointer so anything that happens on the stack
	// will clobber the test pattern created by the caller
	ADD	$(1024 * 8), R1

	// Ask signaller to setgid
	MOVW	$1, R3
	SYNC
	MOVW	R3, ·Baton(SB)

//   - rounds 60-79 are type 4 and do not load data (ROUND4 macro).
//
// Each round loads or shuffles the data, then computes a per-round
// function of b, c, d, and then mixes the result into and rotates the
// five registers a, b, c, d, e holding the intermediate results.
//
// The register rotation is implemented by rotating the arguments to
// the round macros instead of by explicit move instructions.

// Register definitions

	AESDEC X1, X0
	MOVUPS 128(AX), X1
	AESDEC X1, X0
	MOVUPS 144(AX), X1
	AESDECLAST X1, X0
	MOVUPS X0, 0(DX)
	RET

// func expandKeyAsm(nr int, key *byte, enc, dec *uint32) {
// Note that round keys are stored in uint128 format, not uint32
TEXT ·expandKeyAsm(SB),NOSPLIT,$0
	MOVQ nr+0(FP), CX
	MOVQ key+8(FP), AX
	MOVQ enc+16(FP), BX
	MOVQ dec+24(FP), DX
	MOVUPS (AX), X0
	// enc
	MOVUPS X0, (BX)