[picodrive.git] / misc_amips.s

# vim:filetype=mips

# Some misc routines for Allegrex MIPS
# (c) Copyright 2007, Grazvydas "notaz" Ignotas
# All Rights Reserved

.set noreorder
.set noat

.text
.align 4

.globl memset32 # int *dest, int c, int count

memset32:
ms32_aloop:
    andi    $t0, $a0, 0x3f
    beqz    $t0, ms32_bloop_prep
    nop
    sw      $a1, 0($a0)
    addiu   $a2, -1
    beqz    $a2, ms32_return
    addiu   $a0, 4
    j       ms32_aloop
    nop

ms32_bloop_prep:
    srl     $t0, $a2, 4    # we will do 64 bytes per iteration (cache line)
    beqz    $t0, ms32_bloop_end

ms32_bloop:
    addiu   $t0, -1
    cache   0x18, ($a0)    # create dirty exclusive
    sw      $a1, 0x00($a0)
    sw      $a1, 0x04($a0)
    sw      $a1, 0x08($a0)
    sw      $a1, 0x0c($a0)
    sw      $a1, 0x10($a0)
    sw      $a1, 0x14($a0)
    sw      $a1, 0x18($a0)
    sw      $a1, 0x1c($a0)
    sw      $a1, 0x20($a0)
    sw      $a1, 0x24($a0)
    sw      $a1, 0x28($a0)
    sw      $a1, 0x2c($a0)
    sw      $a1, 0x30($a0)
    sw      $a1, 0x34($a0)
    sw      $a1, 0x38($a0)
    sw      $a1, 0x3c($a0)
    bnez    $t0, ms32_bloop
    addiu   $a0, 0x40

ms32_bloop_end:
    andi    $a2, $a2, 0x0f
    beqz    $a2, ms32_return

ms32_cloop:
    addiu   $a2, -1
    sw      $a1, 0($a0)
    bnez    $a2, ms32_cloop
    addiu   $a0, 4

ms32_return:
    jr      $ra
    nop


.globl memset32_uncached # int *dest, int c, int count

memset32_uncached:
    srl     $t0, $a2, 3    # we will do 32 bytes per iteration
    beqz    $t0, ms32u_bloop_end

ms32u_bloop:
    addiu   $t0, -1
    sw      $a1, 0x00($a0)
    sw      $a1, 0x04($a0)
    sw      $a1, 0x08($a0)
    sw      $a1, 0x0c($a0)
    sw      $a1, 0x10($a0)
    sw      $a1, 0x14($a0)
    sw      $a1, 0x18($a0)
    sw      $a1, 0x1c($a0)
    bnez    $t0, ms32u_bloop
    addiu   $a0, 0x20

ms32u_bloop_end:
    andi    $a2, $a2, 0x0f
    beqz    $a2, ms32u_return

ms32u_cloop:
    addiu   $a2, -1
    sw      $a1, 0($a0)
    bnez    $a2, ms32u_cloop
    addiu   $a0, 4

ms32u_return:
    jr      $ra
    nop


.globl memcpy32 # int *dest, int *src, int count

memcpy32:
mc32_aloop:
    andi    $t0, $a0, 0x3f
    beqz    $t0, mc32_bloop_prep
    nop
    lw      $t1, 0($a1)
    addiu   $a2, -1
    sw      $t1, 0($a0)
    beqz    $a2, mc32_return
    addiu   $a0, 4
    j       mc32_aloop
    addiu   $a1, 4

mc32_bloop_prep:
    srl     $t0, $a2, 4    # we will do 64 bytes per iteration (cache line)
    beqz    $t0, mc32_bloop_end

mc32_bloop:
    addiu   $t0, -1
    cache   0x18, ($a0)    # create dirty exclusive
    lw      $t2, 0x00($a1)
    lw      $t3, 0x04($a1)
    lw      $t4, 0x08($a1)
    lw      $t5, 0x0c($a1)
    lw      $t6, 0x10($a1)
    lw      $t7, 0x14($a1)
    lw      $t8, 0x18($a1)
    lw      $t9, 0x1c($a1)
    sw      $t2, 0x00($a0)
    sw      $t3, 0x04($a0)
    sw      $t4, 0x08($a0)
    sw      $t5, 0x0c($a0)
    sw      $t6, 0x10($a0)
    sw      $t7, 0x14($a0)
    sw      $t8, 0x18($a0)
    sw      $t9, 0x1c($a0)
    lw      $t2, 0x20($a1)
    lw      $t3, 0x24($a1)
    lw      $t4, 0x28($a1)
    lw      $t5, 0x2c($a1)
    lw      $t6, 0x30($a1)
    lw      $t7, 0x34($a1)
    lw      $t8, 0x38($a1)
    lw      $t9, 0x3c($a1)
    sw      $t2, 0x20($a0)
    sw      $t3, 0x24($a0)
    sw      $t4, 0x28($a0)
    sw      $t5, 0x2c($a0)
    sw      $t6, 0x30($a0)
    sw      $t7, 0x34($a0)
    sw      $t8, 0x38($a0)
    sw      $t9, 0x3c($a0)
    addiu   $a0, 0x40
    bnez    $t0, mc32_bloop
    addiu   $a1, 0x40

mc32_bloop_end:
    andi    $a2, $a2, 0x0f
    beqz    $a2, mc32_return

mc32_cloop:
    lw      $t1, 0($a1)
    addiu   $a2, -1
    addiu   $a1, 4
    sw      $t1, 0($a0)
    bnez    $a2, mc32_cloop
    addiu   $a0, 4

mc32_return:
    jr      $ra
    nop
Commit	Line	Data
	1	# vim:filetype=mips
	2
	3	# Some misc routines for Allegrex MIPS
	4	# (c) Copyright 2007, Grazvydas "notaz" Ignotas
	5	# All Rights Reserved
	6
	7	.set noreorder
	8	.set noat
	9
	10	.text
	11	.align 4
	12
	13	.globl memset32 # int *dest, int c, int count
	14
	15	memset32:
	16	ms32_aloop:
	17	andi $t0, $a0, 0x3f
	18	beqz $t0, ms32_bloop_prep
	19	nop
	20	sw $a1, 0($a0)
	21	addiu $a2, -1
	22	beqz $a2, ms32_return
	23	addiu $a0, 4
	24	j ms32_aloop
	25	nop
	26
	27	ms32_bloop_prep:
	28	srl $t0, $a2, 4 # we will do 64 bytes per iteration (cache line)
	29	beqz $t0, ms32_bloop_end
	30
	31	ms32_bloop:
	32	addiu $t0, -1
	33	cache 0x18, ($a0) # create dirty exclusive
	34	sw $a1, 0x00($a0)
	35	sw $a1, 0x04($a0)
	36	sw $a1, 0x08($a0)
	37	sw $a1, 0x0c($a0)
	38	sw $a1, 0x10($a0)
	39	sw $a1, 0x14($a0)
	40	sw $a1, 0x18($a0)
	41	sw $a1, 0x1c($a0)
	42	sw $a1, 0x20($a0)
	43	sw $a1, 0x24($a0)
	44	sw $a1, 0x28($a0)
	45	sw $a1, 0x2c($a0)
	46	sw $a1, 0x30($a0)
	47	sw $a1, 0x34($a0)
	48	sw $a1, 0x38($a0)
	49	sw $a1, 0x3c($a0)
	50	bnez $t0, ms32_bloop
	51	addiu $a0, 0x40
	52
	53	ms32_bloop_end:
	54	andi $a2, $a2, 0x0f
	55	beqz $a2, ms32_return
	56
	57	ms32_cloop:
	58	addiu $a2, -1
	59	sw $a1, 0($a0)
	60	bnez $a2, ms32_cloop
	61	addiu $a0, 4
	62
	63	ms32_return:
	64	jr $ra
	65	nop
	66
	67
	68	.globl memset32_uncached # int *dest, int c, int count
	69
	70	memset32_uncached:
	71	srl $t0, $a2, 3 # we will do 32 bytes per iteration
	72	beqz $t0, ms32u_bloop_end
	73
	74	ms32u_bloop:
	75	addiu $t0, -1
	76	sw $a1, 0x00($a0)
	77	sw $a1, 0x04($a0)
	78	sw $a1, 0x08($a0)
	79	sw $a1, 0x0c($a0)
	80	sw $a1, 0x10($a0)
	81	sw $a1, 0x14($a0)
	82	sw $a1, 0x18($a0)
	83	sw $a1, 0x1c($a0)
	84	bnez $t0, ms32u_bloop
	85	addiu $a0, 0x20
	86
	87	ms32u_bloop_end:
	88	andi $a2, $a2, 0x0f
	89	beqz $a2, ms32u_return
	90
	91	ms32u_cloop:
	92	addiu $a2, -1
	93	sw $a1, 0($a0)
	94	bnez $a2, ms32u_cloop
	95	addiu $a0, 4
	96
	97	ms32u_return:
	98	jr $ra
	99	nop
	100
	101
	102	.globl memcpy32 # int dest, int src, int count
	103
	104	memcpy32:
	105	mc32_aloop:
	106	andi $t0, $a0, 0x3f
	107	beqz $t0, mc32_bloop_prep
	108	nop
	109	lw $t1, 0($a1)
	110	addiu $a2, -1
	111	sw $t1, 0($a0)
	112	beqz $a2, mc32_return
	113	addiu $a0, 4
	114	j mc32_aloop
	115	addiu $a1, 4
	116
	117	mc32_bloop_prep:
	118	srl $t0, $a2, 4 # we will do 64 bytes per iteration (cache line)
	119	beqz $t0, mc32_bloop_end
	120
	121	mc32_bloop:
	122	addiu $t0, -1
	123	cache 0x18, ($a0) # create dirty exclusive
	124	lw $t2, 0x00($a1)
	125	lw $t3, 0x04($a1)
	126	lw $t4, 0x08($a1)
	127	lw $t5, 0x0c($a1)
	128	lw $t6, 0x10($a1)
	129	lw $t7, 0x14($a1)
	130	lw $t8, 0x18($a1)
	131	lw $t9, 0x1c($a1)
	132	sw $t2, 0x00($a0)
	133	sw $t3, 0x04($a0)
	134	sw $t4, 0x08($a0)
	135	sw $t5, 0x0c($a0)
	136	sw $t6, 0x10($a0)
	137	sw $t7, 0x14($a0)
	138	sw $t8, 0x18($a0)
	139	sw $t9, 0x1c($a0)
	140	lw $t2, 0x20($a1)
	141	lw $t3, 0x24($a1)
	142	lw $t4, 0x28($a1)
	143	lw $t5, 0x2c($a1)
	144	lw $t6, 0x30($a1)
	145	lw $t7, 0x34($a1)
	146	lw $t8, 0x38($a1)
	147	lw $t9, 0x3c($a1)
	148	sw $t2, 0x20($a0)
	149	sw $t3, 0x24($a0)
	150	sw $t4, 0x28($a0)
	151	sw $t5, 0x2c($a0)
	152	sw $t6, 0x30($a0)
	153	sw $t7, 0x34($a0)
	154	sw $t8, 0x38($a0)
	155	sw $t9, 0x3c($a0)
	156	addiu $a0, 0x40
	157	bnez $t0, mc32_bloop
	158	addiu $a1, 0x40
	159
	160	mc32_bloop_end:
	161	andi $a2, $a2, 0x0f
	162	beqz $a2, mc32_return
	163
	164	mc32_cloop:
	165	lw $t1, 0($a1)
	166	addiu $a2, -1
	167	addiu $a1, 4
	168	sw $t1, 0($a0)
	169	bnez $a2, mc32_cloop
	170	addiu $a0, 4
	171
	172	mc32_return:
	173	jr $ra
	174	nop
	175