#include "funcdata.h"
#include "textflag.h"

// If !iscgo, this is a no-op.
//
// NOTE: mcall() assumes this clobbers only R23 (REGTMP).
TEXT runtime·save_g(SB),NOSPLIT|NOFRAME,$0-0
	MOVB	runtime·iscgo(SB), R23
	BEQ	R23, nocgo

	MOVV	R3, R23	// save R3
	MOVV	g, runtime·tls_g(SB) // TLS relocation clobbers R3
	MOVV	R23, R3	// restore R3

nocgo:
	RET

TEXT runtime·load_g(SB),NOSPLIT|NOFRAME,$0-0

#include "funcdata.h"
#include "textflag.h"

// If !iscgo, this is a no-op.
// NOTE: gogo assumes load_g only clobers g (R30) and REGTMP (R23)
TEXT runtime·save_g(SB),NOSPLIT|NOFRAME,$0-0
	MOVB	runtime·iscgo(SB), R23
	BEQ	R23, nocgo

	MOVW	R3, R23
	MOVW	g, runtime·tls_g(SB) // TLS relocation clobbers R3
	MOVW	R23, R3

nocgo:
	RET

TEXT runtime·load_g(SB),NOSPLIT|NOFRAME,$0-0
	MOVW	runtime·tls_g(SB), g // TLS relocation clobbers R3

 *
 * Calling into the gc tool chain, where all registers are caller save.
 * Called from standard lp64d ABI, where $r1, $r3, $r23-$r30, and $f24-$f31
 * are callee-save, so they must be saved explicitly, along with $r1 (LR).
 */
.globl crosscall1
crosscall1:
	addi.d	$r3, $r3, -160
	st.d	$r1, $r3, 0
	st.d	$r23, $r3, 8
	st.d	$r24, $r3, 16
	st.d	$r25, $r3, 24
	st.d	$r26, $r3, 32
	st.d	$r27, $r3, 40
	st.d	$r28, $r3, 48

	// Find the offset to those bytes
	SUB R5, R22, R22
	SLD $3, R22

	// Shift to get only the bytes in msg
	SRD R22, R17, R17

	// Put 1 at high end
	MOVD $1, R23
	SLD  $3, R21
	SLD  R21, R23, R23
	OR   R23, R17, R17

	// Remainder is 8
	MOVD $8, R5

just1:
	CMP R5, $8
	BLT less8

	// Exactly 8
	MOVD (R4), R16

	CMP R17, $0

TEXT	asancall<>(SB), NOSPLIT, $0-0
	MOVV	R3, R23         // callee-saved
	BEQ	g, g0stack      // no g, still on a system stack
	MOVV	g_m(g), R14
	MOVV	m_g0(R14), R15
	BEQ	R15, g, g0stack

	MOVV	(g_sched+gobuf_sp)(R15), R9
	MOVV	R9, R3

g0stack:
	JAL	(FARG)
	MOVV	R23, R3

	STP (R6, R7), 56(RSP)
	STP (R8, R9), 72(RSP)
	STP (R10, R11), 88(RSP)
	STP (R12, R13), 104(RSP)
	STP (R14, R15), 120(RSP)
	STP (R16, R17), 136(RSP)
	STP (R19, R20), 152(RSP)
	STP (R21, R22), 168(RSP)
	STP (R23, R24), 184(RSP)
	STP (R25, R26), 200(RSP)
	MOVD NZCV, R0
	MOVD R0, 216(RSP)
	MOVD FPSR, R0
	MOVD R0, 224(RSP)
	FSTPD (F0, F1), 232(RSP)
	FSTPD (F2, F3), 248(RSP)
	FSTPD (F4, F5), 264(RSP)

	STP	(R19, R20), ((offset)+0*8)(RSP) \
	STP	(R21, R22), ((offset)+2*8)(RSP) \
	STP	(R23, R24), ((offset)+4*8)(RSP) \
	STP	(R25, R26), ((offset)+6*8)(RSP) \
	STP	(R27, g), ((offset)+8*8)(RSP)

#define RESTORE_R19_TO_R28(offset) \
	LDP	((offset)+0*8)(RSP), (R19, R20) \
	LDP	((offset)+2*8)(RSP), (R21, R22) \
	LDP	((offset)+4*8)(RSP), (R23, R24) \
	LDP	((offset)+6*8)(RSP), (R25, R26) \
	LDP	((offset)+8*8)(RSP), (R27, g) /* R28 */

	print("r18  ", hex(c.r18()), "\t")
	print("r19  ", hex(c.r19()), "\n")
	print("r20  ", hex(c.r20()), "\t")
	print("r21  ", hex(c.r21()), "\n")
	print("r22  ", hex(c.r22()), "\t")
	print("r23  ", hex(c.r23()), "\n")
	print("r24  ", hex(c.r24()), "\t")
	print("r25  ", hex(c.r25()), "\n")
	print("r26  ", hex(c.r26()), "\t")
	print("r27  ", hex(c.r27()), "\n")
	print("r28  ", hex(c.r28()), "\t")

TEXT	msancall<>(SB), NOSPLIT, $0-0
	MOVV	R3, R23         // callee-saved
	BEQ	g, g0stack      // no g, still on a system stack
	MOVV	g_m(g), R14
	MOVV	m_g0(R14), R15
	BEQ	R15, g, g0stack

	MOVV	(g_sched+gobuf_sp)(R15), R9
	MOVV	R9, R3

g0stack:
	JAL	(FARG)
	MOVV	R23, R3

//
// SAVE_F24_TO_F31(offset) saves F24 ~ F31 to the stack space
// of ((offset)+0*8)(R3) ~ ((offset)+7*8)(R3).
//
// Note: g is R22

#define SAVE_R22_TO_R31(offset)	\
	MOVV	g,   ((offset)+(0*8))(R3)	\
	MOVV	R23, ((offset)+(1*8))(R3)	\
	MOVV	R24, ((offset)+(2*8))(R3)	\
	MOVV	R25, ((offset)+(3*8))(R3)	\
	MOVV	R26, ((offset)+(4*8))(R3)	\
	MOVV	R27, ((offset)+(5*8))(R3)	\
	MOVV	R28, ((offset)+(6*8))(R3)	\