/*
 * arch/alpha/kernel/entry.S
 *
 * Kernel entry-points.
 */

#include <asm/asm-offsets.h>
#include <asm/thread_info.h>
#include <asm/pal.h>
#include <asm/errno.h>
#include <asm/unistd.h>

	.text
	.set noat

/* Stack offsets.  */
#define SP_OFF			184
#define SWITCH_STACK_SIZE	320

/*
 * This defines the normal kernel pt-regs layout.
 *
 * regs 9-15 preserved by C code
 * regs 16-18 saved by PAL-code
 * regs 29-30 saved and set up by PAL-code
 * JRP - Save regs 16-18 in a special area of the stack, so that
 * the palcode-provided values are available to the signal handler.
 */

#define SAVE_ALL			\
	subq	$sp, SP_OFF, $sp;	\
	stq	$0, 0($sp);		\
	stq	$1, 8($sp);		\
	stq	$2, 16($sp);		\
	stq	$3, 24($sp);		\
	stq	$4, 32($sp);		\
	stq	$28, 144($sp);		\
	lda	$2, alpha_mv;		\
	stq	$5, 40($sp);		\
	stq	$6, 48($sp);		\
	stq	$7, 56($sp);		\
	stq	$8, 64($sp);		\
	stq	$19, 72($sp);		\
	stq	$20, 80($sp);		\
	stq	$21, 88($sp);		\
	ldq	$2, HAE_CACHE($2);	\
	stq	$22, 96($sp);		\
	stq	$23, 104($sp);		\
	stq	$24, 112($sp);		\
	stq	$25, 120($sp);		\
	stq	$26, 128($sp);		\
	stq	$27, 136($sp);		\
	stq	$2, 152($sp);		\
	stq	$16, 160($sp);		\
	stq	$17, 168($sp);		\
	stq	$18, 176($sp)

#define RESTORE_ALL			\
	lda	$19, alpha_mv;		\
	ldq	$0, 0($sp);		\
	ldq	$1, 8($sp);		\
	ldq	$2, 16($sp);		\
	ldq	$3, 24($sp);		\
	ldq	$21, 152($sp);		\
	ldq	$20, HAE_CACHE($19);	\
	ldq	$4, 32($sp);		\
	ldq	$5, 40($sp);		\
	ldq	$6, 48($sp);		\
	ldq	$7, 56($sp);		\
	subq	$20, $21, $20;		\
	ldq	$8, 64($sp);		\
	beq	$20, 99f;		\
	ldq	$20, HAE_REG($19);	\
	stq	$21, HAE_CACHE($19);	\
	stq	$21, 0($20);		\
99:;					\
	ldq	$19, 72($sp);		\
	ldq	$20, 80($sp);		\
	ldq	$21, 88($sp);		\
	ldq	$22, 96($sp);		\
	ldq	$23, 104($sp);		\
	ldq	$24, 112($sp);		\
	ldq	$25, 120($sp);		\
	ldq	$26, 128($sp);		\
	ldq	$27, 136($sp);		\
	ldq	$28, 144($sp);		\
	addq	$sp, SP_OFF, $sp

/*
 * Non-syscall kernel entry points.
 */

	.align	4
	.globl	entInt
	.ent	entInt
entInt:
	SAVE_ALL
	lda	$8, 0x3fff
	lda	$26, ret_from_sys_call
	bic	$sp, $8, $8
	mov	$sp, $19
	jsr	$31, do_entInt
.end entInt

	.align	4
	.globl	entArith
	.ent	entArith
entArith:
	SAVE_ALL
	lda	$8, 0x3fff
	lda	$26, ret_from_sys_call
	bic	$sp, $8, $8
	mov	$sp, $18
	jsr	$31, do_entArith
.end entArith

	.align	4
	.globl	entMM
	.ent	entMM
entMM:
	SAVE_ALL
/* save $9 - $15 so the inline exception code can manipulate them.  */
	subq	$sp, 56, $sp
	stq	$9, 0($sp)
	stq	$10, 8($sp)
	stq	$11, 16($sp)
	stq	$12, 24($sp)
	stq	$13, 32($sp)
	stq	$14, 40($sp)
	stq	$15, 48($sp)
	addq	$sp, 56, $19
/* handle the fault */
	lda	$8, 0x3fff
	bic	$sp, $8, $8
	jsr	$26, do_page_fault
/* reload the registers after the exception code played.  */
	ldq	$9, 0($sp)
	ldq	$10, 8($sp)
	ldq	$11, 16($sp)
	ldq	$12, 24($sp)
	ldq	$13, 32($sp)
	ldq	$14, 40($sp)
	ldq	$15, 48($sp)
	addq	$sp, 56, $sp
/* finish up the syscall as normal.  */
	br	ret_from_sys_call
.end entMM

	.align	4
	.globl	entIF
	.ent	entIF
entIF:
	SAVE_ALL
	lda	$8, 0x3fff
	lda	$26, ret_from_sys_call
	bic	$sp, $8, $8
	mov	$sp, $17
	jsr	$31, do_entIF
.end entIF

	.align	4
	.globl	entUna
	.ent	entUna
entUna:
	lda	$sp, -256($sp)
	stq	$0, 0($sp)
	ldq	$0, 256($sp)	/* get PS */
	stq	$1, 8($sp)
	stq	$2, 16($sp)
	stq	$3, 24($sp)
	and	$0, 8, $0		/* user mode? */
	stq	$4, 32($sp)
	bne	$0, entUnaUser	/* yup -> do user-level unaligned fault */
	stq	$5, 40($sp)
	stq	$6, 48($sp)
	stq	$7, 56($sp)
	stq	$8, 64($sp)
	stq	$9, 72($sp)
	stq	$10, 80($sp)
	stq	$11, 88($sp)
	stq	$12, 96($sp)
	stq	$13, 104($sp)
	stq	$14, 112($sp)
	stq	$15, 120($sp)
	/* 16-18 PAL-saved */
	stq	$19, 152($sp)
	stq	$20, 160($sp)
	stq	$21, 168($sp)
	stq	$22, 176($sp)
	stq	$23, 184($sp)
	stq	$24, 192($sp)
	stq	$25, 200($sp)
	stq	$26, 208($sp)
	stq	$27, 216($sp)
	stq	$28, 224($sp)
	mov	$sp, $19
	stq	$gp, 232($sp)
	lda	$8, 0x3fff
	stq	$31, 248($sp)
	bic	$sp, $8, $8
	jsr	$26, do_entUna
	ldq	$0, 0($sp)
	ldq	$1, 8($sp)
	ldq	$2, 16($sp)
	ldq	$3, 24($sp)
	ldq	$4, 32($sp)
	ldq	$5, 40($sp)
	ldq	$6, 48($sp)
	ldq	$7, 56($sp)
	ldq	$8, 64($sp)
	ldq	$9, 72($sp)
	ldq	$10, 80($sp)
	ldq	$11, 88($sp)
	ldq	$12, 96($sp)
	ldq	$13, 104($sp)
	ldq	$14, 112($sp)
	ldq	$15, 120($sp)
	/* 16-18 PAL-saved */
	ldq	$19, 152($sp)
	ldq	$20, 160($sp)
	ldq	$21, 168($sp)
	ldq	$22, 176($sp)
	ldq	$23, 184($sp)
	ldq	$24, 192($sp)
	ldq	$25, 200($sp)
	ldq	$26, 208($sp)
	ldq	$27, 216($sp)
	ldq	$28, 224($sp)
	ldq	$gp, 232($sp)
	lda	$sp, 256($sp)
	call_pal PAL_rti
.end entUna

	.align	4
	.ent	entUnaUser
entUnaUser:
	ldq	$0, 0($sp)	/* restore original $0 */
	lda	$sp, 256($sp)	/* pop entUna's stack frame */
	SAVE_ALL		/* setup normal kernel stack */
	lda	$sp, -56($sp)
	stq	$9, 0($sp)
	stq	$10, 8($sp)
	stq	$11, 16($sp)
	stq	$12, 24($sp)
	stq	$13, 32($sp)
	stq	$14, 40($sp)
	stq	$15, 48($sp)
	lda	$8, 0x3fff
	addq	$sp, 56, $19
	bic	$sp, $8, $8
	jsr	$26, do_entUnaUser
	ldq	$9, 0($sp)
	ldq	$10, 8($sp)
	ldq	$11, 16($sp)
	ldq	$12, 24($sp)
	ldq	$13, 32($sp)
	ldq	$14, 40($sp)
	ldq	$15, 48($sp)
	lda	$sp, 56($sp)
	br	ret_from_sys_call
.end entUnaUser

	.align	4
	.globl	entDbg
	.ent	entDbg
entDbg:
	SAVE_ALL
	lda	$8, 0x3fff
	lda	$26, ret_from_sys_call
	bic	$sp, $8, $8
	mov	$sp, $16
	jsr	$31, do_entDbg
.end entDbg

/*
 * The system call entry point is special.  Most importantly, it looks
 * like a function call to userspace as far as clobbered registers.  We
 * do preserve the argument registers (for syscall restarts) and $26
 * (for leaf syscall functions).
 *
 * So much for theory.  We don't take advantage of this yet.
 *
 * Note that a0-a2 are not saved by PALcode as with the other entry points.
 */

	.align	4
	.globl	entSys
	.globl	ret_from_sys_call
	.ent	entSys
entSys:
	SAVE_ALL
	lda	$8, 0x3fff
	bic	$sp, $8, $8
	lda	$4, NR_SYSCALLS($31)
	stq	$16, SP_OFF+24($sp)
	lda	$5, sys_call_table
	lda	$27, sys_ni_syscall
	cmpult	$0, $4, $4
	ldl	$3, TI_FLAGS($8)
	stq	$17, SP_OFF+32($sp)
	s8addq	$0, $5, $5
	stq	$18, SP_OFF+40($sp)
	blbs	$3, strace
	beq	$4, 1f
	ldq	$27, 0($5)
1:	jsr	$26, ($27), alpha_ni_syscall
	ldgp	$gp, 0($26)
	blt	$0, $syscall_error	/* the call failed */
	stq	$0, 0($sp)
	stq	$31, 72($sp)		/* a3=0 => no error */

	.align	4
ret_from_sys_call:
	cmovne	$26, 0, $19		/* $19 = 0 => non-restartable */
	ldq	$0, SP_OFF($sp)
	and	$0, 8, $0
	beq	$0, ret_to_kernel
ret_to_user:
	/* Make sure need_resched and sigpending don't change between
		sampling and the rti.  */
	lda	$16, 7
	call_pal PAL_swpipl
	ldl	$5, TI_FLAGS($8)
	and	$5, _TIF_WORK_MASK, $2
	bne	$2, work_pending
restore_all:
	RESTORE_ALL
	call_pal PAL_rti

ret_to_kernel:
	lda	$16, 7
	call_pal PAL_swpipl
	br restore_all

	.align 3
$syscall_error:
	/*
	 * Some system calls (e.g., ptrace) can return arbitrary
	 * values which might normally be mistaken as error numbers.
	 * Those functions must zero $0 (v0) directly in the stack
	 * frame to indicate that a negative return value wasn't an
	 * error number..
	 */
	ldq	$19, 0($sp)	/* old syscall nr (zero if success) */
	beq	$19, $ret_success

	ldq	$20, 72($sp)	/* .. and this a3 */
	subq	$31, $0, $0	/* with error in v0 */
	addq	$31, 1, $1	/* set a3 for errno return */
	stq	$0, 0($sp)
	mov	$31, $26	/* tell "ret_from_sys_call" we can restart */
	stq	$1, 72($sp)	/* a3 for return */
	br	ret_from_sys_call

$ret_success:
	stq	$0, 0($sp)
	stq	$31, 72($sp)	/* a3=0 => no error */
	br	ret_from_sys_call
.end entSys

/*
 * Do all cleanup when returning from all interrupts and system calls.
 *
 * Arguments:
 *       $5: TI_FLAGS.
 *       $8: current.
 *      $19: The old syscall number, or zero if this is not a return
 *           from a syscall that errored and is possibly restartable.
 *      $20: The old a3 value
 */

	.align	4
	.ent	work_pending
work_pending:
	and	$5, _TIF_NEED_RESCHED, $2
	beq	$2, $work_notifysig

$work_resched:
	subq	$sp, 16, $sp
	stq	$19, 0($sp)              /* save syscall nr */
	stq	$20, 8($sp)              /* and error indication (a3) */
	jsr	$26, schedule
	ldq	$19, 0($sp)
	ldq	$20, 8($sp)
	addq	$sp, 16, $sp
	/* Make sure need_resched and sigpending don't change between
		sampling and the rti.  */
	lda	$16, 7
	call_pal PAL_swpipl
	ldl	$5, TI_FLAGS($8)
	and	$5, _TIF_WORK_MASK, $2
	beq	$2, restore_all
	and	$5, _TIF_NEED_RESCHED, $2
	bne	$2, $work_resched

$work_notifysig:
	mov	$sp, $16
	bsr	$1, do_switch_stack
	mov	$sp, $17
	mov	$5, $18
	mov	$19, $9		/* save old syscall number */
	mov	$20, $10	/* save old a3 */
	and	$5, _TIF_SIGPENDING, $2
	cmovne	$2, 0, $9	/* we don't want double syscall restarts */
	jsr	$26, do_notify_resume
	mov	$9, $19
	mov	$10, $20
	bsr	$1, undo_switch_stack
	br	ret_to_user
.end work_pending

/*
 * PTRACE syscall handler
 */

	.align	4
	.ent	strace
strace:
	/* set up signal stack, call syscall_trace */
	bsr	$1, do_switch_stack
	jsr	$26, syscall_trace
	bsr	$1, undo_switch_stack

	/* get the system call number and the arguments back.. */
	ldq	$0, 0($sp)
	ldq	$16, SP_OFF+24($sp)
	ldq	$17, SP_OFF+32($sp)
	ldq	$18, SP_OFF+40($sp)
	ldq	$19, 72($sp)
	ldq	$20, 80($sp)
	ldq	$21, 88($sp)

	/* get the system call pointer.. */
	lda	$1, NR_SYSCALLS($31)
	lda	$2, sys_call_table
	lda	$27, alpha_ni_syscall
	cmpult	$0, $1, $1
	s8addq	$0, $2, $2
	beq	$1, 1f
	ldq	$27, 0($2)
1:	jsr	$26, ($27), sys_gettimeofday
ret_from_straced:
	ldgp	$gp, 0($26)

	/* check return.. */
	blt	$0, $strace_error	/* the call failed */
	stq	$31, 72($sp)		/* a3=0 => no error */
$strace_success:
	stq	$0, 0($sp)		/* save return value */

	bsr	$1, do_switch_stack
	jsr	$26, syscall_trace
	bsr	$1, undo_switch_stack
	br	$31, ret_from_sys_call

	.align	3
$strace_error:
	ldq	$19, 0($sp)	/* old syscall nr (zero if success) */
	beq	$19, $strace_success
	ldq	$20, 72($sp)	/* .. and this a3 */

	subq	$31, $0, $0	/* with error in v0 */
	addq	$31, 1, $1	/* set a3 for errno return */
	stq	$0, 0($sp)
	stq	$1, 72($sp)	/* a3 for return */

	bsr	$1, do_switch_stack
	mov	$19, $9		/* save old syscall number */
	mov	$20, $10	/* save old a3 */
	jsr	$26, syscall_trace
	mov	$9, $19
	mov	$10, $20
	bsr	$1, undo_switch_stack

	mov	$31, $26	/* tell "ret_from_sys_call" we can restart */
	br	ret_from_sys_call
.end strace

/*
 * Save and restore the switch stack -- aka the balance of the user context.
 */

	.align	4
	.ent	do_switch_stack
do_switch_stack:
	lda	$sp, -SWITCH_STACK_SIZE($sp)
	stq	$9, 0($sp)
	stq	$10, 8($sp)
	stq	$11, 16($sp)
	stq	$12, 24($sp)
	stq	$13, 32($sp)
	stq	$14, 40($sp)
	stq	$15, 48($sp)
	stq	$26, 56($sp)
	stt	$f0, 64($sp)
	stt	$f1, 72($sp)
	stt	$f2, 80($sp)
	stt	$f3, 88($sp)
	stt	$f4, 96($sp)
	stt	$f5, 104($sp)
	stt	$f6, 112($sp)
	stt	$f7, 120($sp)
	stt	$f8, 128($sp)
	stt	$f9, 136($sp)
	stt	$f10, 144($sp)
	stt	$f11, 152($sp)
	stt	$f12, 160($sp)
	stt	$f13, 168($sp)
	stt	$f14, 176($sp)
	stt	$f15, 184($sp)
	stt	$f16, 192($sp)
	stt	$f17, 200($sp)
	stt	$f18, 208($sp)
	stt	$f19, 216($sp)
	stt	$f20, 224($sp)
	stt	$f21, 232($sp)
	stt	$f22, 240($sp)
	stt	$f23, 248($sp)
	stt	$f24, 256($sp)
	stt	$f25, 264($sp)
	stt	$f26, 272($sp)
	stt	$f27, 280($sp)
	mf_fpcr	$f0		# get fpcr
	stt	$f28, 288($sp)
	stt	$f29, 296($sp)
	stt	$f30, 304($sp)
	stt	$f0, 312($sp)	# save fpcr in slot of $f31
	ldt	$f0, 64($sp)	# dont let "do_switch_stack" change fp state.
	ret	$31, ($1), 1
.end do_switch_stack

	.align	4
	.ent	undo_switch_stack
undo_switch_stack:
	ldq	$9, 0($sp)
	ldq	$10, 8($sp)
	ldq	$11, 16($sp)
	ldq	$12, 24($sp)
	ldq	$13, 32($sp)
	ldq	$14, 40($sp)
	ldq	$15, 48($sp)
	ldq	$26, 56($sp)
	ldt	$f30, 312($sp)	# get saved fpcr
	ldt	$f0, 64($sp)
	ldt	$f1, 72($sp)
	ldt	$f2, 80($sp)
	ldt	$f3, 88($sp)
	mt_fpcr	$f30		# install saved fpcr
	ldt	$f4, 96($sp)
	ldt	$f5, 104($sp)
	ldt	$f6, 112($sp)
	ldt	$f7, 120($sp)
	ldt	$f8, 128($sp)
	ldt	$f9, 136($sp)
	ldt	$f10, 144($sp)
	ldt	$f11, 152($sp)
	ldt	$f12, 160($sp)
	ldt	$f13, 168($sp)
	ldt	$f14, 176($sp)
	ldt	$f15, 184($sp)
	ldt	$f16, 192($sp)
	ldt	$f17, 200($sp)
	ldt	$f18, 208($sp)
	ldt	$f19, 216($sp)
	ldt	$f20, 224($sp)
	ldt	$f21, 232($sp)
	ldt	$f22, 240($sp)
	ldt	$f23, 248($sp)
	ldt	$f24, 256($sp)
	ldt	$f25, 264($sp)
	ldt	$f26, 272($sp)
	ldt	$f27, 280($sp)
	ldt	$f28, 288($sp)
	ldt	$f29, 296($sp)
	ldt	$f30, 304($sp)
	lda	$sp, SWITCH_STACK_SIZE($sp)
	ret	$31, ($1), 1
.end undo_switch_stack

/*
 * The meat of the context switch code.
 */

	.align	4
	.globl	alpha_switch_to
	.ent	alpha_switch_to
alpha_switch_to:
	.prologue 0
	bsr	$1, do_switch_stack
	call_pal PAL_swpctx
	lda	$8, 0x3fff
	bsr	$1, undo_switch_stack
	bic	$sp, $8, $8
	mov	$17, $0
	ret
.end alpha_switch_to

/*
 * New processes begin life here.
 */

	.globl	ret_from_fork
	.align	4
	.ent	ret_from_fork
ret_from_fork:
	lda	$26, ret_from_sys_call
	mov	$17, $16
	jmp	$31, schedule_tail
.end ret_from_fork

/*
 * kernel_thread(fn, arg, clone_flags)
 */
	.align 4
	.globl	kernel_thread
	.ent	kernel_thread
kernel_thread:
	/* We can be called from a module.  */
	ldgp	$gp, 0($27)
	.prologue 1
	subq	$sp, SP_OFF+6*8, $sp
	br	$1, 2f		/* load start address */

	/* We've now "returned" from a fake system call.  */
	unop
	blt	$0, 1f		/* error?  */
	ldi	$1, 0x3fff
	beq	$20, 1f		/* parent or child?  */

	bic	$sp, $1, $8	/* in child.  */
	jsr	$26, ($27)
	ldgp	$gp, 0($26)
	mov	$0, $16
	mov	$31, $26
	jmp	$31, sys_exit

1:	ret			/* in parent.  */

	.align 4
2:	/* Fake a system call stack frame, as we can't do system calls
	   from kernel space.  Note that we store FN and ARG as they
	   need to be set up in the child for the call.  Also store $8
	   and $26 for use in the parent.  */
	stq	$31, SP_OFF($sp)	/* ps */
	stq	$1, SP_OFF+8($sp)	/* pc */
	stq	$gp, SP_OFF+16($sp)	/* gp */
	stq	$16, 136($sp)		/* $27; FN for child */
	stq	$17, SP_OFF+24($sp)	/* $16; ARG for child */
	stq	$8, 64($sp)		/* $8 */
	stq	$26, 128($sp)		/* $26 */
	/* Avoid the HAE being gratuitously wrong, to avoid restoring it.  */
	ldq	$2, alpha_mv+HAE_CACHE
	stq	$2, 152($sp)		/* HAE */

	/* Shuffle FLAGS to the front; add CLONE_VM.  */
	ldi	$1, CLONE_VM|CLONE_UNTRACED
	or	$18, $1, $16
	bsr	$26, sys_clone

	/* We don't actually care for a3 success widgetry in the kernel.
	   Not for positive errno values.  */
	stq	$0, 0($sp)		/* $0 */
	br	ret_to_kernel
.end kernel_thread


/*
 * Special system calls.  Most of these are special in that they either
 * have to play switch_stack games or in some way use the pt_regs struct.
 */
	.align	4
	.globl	sys_fork
	.ent	sys_fork
sys_fork:
	.prologue 0
	mov	$sp, $21
	bsr	$1, do_switch_stack
	bis	$31, SIGCHLD, $16
	mov	$31, $17
	mov	$31, $18
	mov	$31, $19
	mov	$31, $20
	jsr	$26, alpha_clone
	bsr	$1, undo_switch_stack
	ret
.end sys_fork

	.align	4
	.globl	sys_clone
	.ent	sys_clone
sys_clone:
	.prologue 0
	mov	$sp, $21
	bsr	$1, do_switch_stack
	/* $16, $17, $18, $19, $20 come from the user.  */
	jsr	$26, alpha_clone
	bsr	$1, undo_switch_stack
	ret
.end sys_clone

	.align	4
	.globl	sys_vfork
	.ent	sys_vfork
sys_vfork:
	.prologue 0
	mov	$sp, $16
	bsr	$1, do_switch_stack
	jsr	$26, alpha_vfork
	bsr	$1, undo_switch_stack
	ret
.end sys_vfork

	.align	4
	.globl	sys_sigreturn
	.ent	sys_sigreturn
sys_sigreturn:
	.prologue 0
	lda	$9, ret_from_straced
	cmpult	$26, $9, $9
	mov	$sp, $17
	lda	$18, -SWITCH_STACK_SIZE($sp)
	lda	$sp, -SWITCH_STACK_SIZE($sp)
	jsr	$26, do_sigreturn
	bne	$9, 1f
	jsr	$26, syscall_trace
1:	br	$1, undo_switch_stack
	br	ret_from_sys_call
.end sys_sigreturn

	.align	4
	.globl	sys_rt_sigreturn
	.ent	sys_rt_sigreturn
sys_rt_sigreturn:
	.prologue 0
	lda	$9, ret_from_straced
	cmpult	$26, $9, $9
	mov	$sp, $17
	lda	$18, -SWITCH_STACK_SIZE($sp)
	lda	$sp, -SWITCH_STACK_SIZE($sp)
	jsr	$26, do_rt_sigreturn
	bne	$9, 1f
	jsr	$26, syscall_trace
1:	br	$1, undo_switch_stack
	br	ret_from_sys_call
.end sys_rt_sigreturn

	.align	4
	.globl	sys_execve
	.ent	sys_execve
sys_execve:
	.prologue 0
	mov	$sp, $19
	jmp	$31, do_sys_execve
.end sys_execve

	.align	4
	.globl	alpha_ni_syscall
	.ent	alpha_ni_syscall
alpha_ni_syscall:
	.prologue 0
	/* Special because it also implements overflow handling via
	   syscall number 0.  And if you recall, zero is a special
	   trigger for "not an error".  Store large non-zero there.  */
	lda	$0, -ENOSYS
	unop
	stq	$0, 0($sp)
	ret
.end alpha_ni_syscall