[pcsx_rearmed.git] / plugins / dfsound / arm_utils.s

/*
 * (C) Gražvydas "notaz" Ignotas, 2011
 *
 * This work is licensed under the terms of any of these licenses
 * (at your option):
 *  - GNU GPL, version 2 or later.
 *  - GNU LGPL, version 2.1 or later.
 * See the COPYING file in the top-level directory.
 */


.text
.align 2


.global mix_chan @ (int start, int count, int lv, int rv)
mix_chan:
    vmov.32     d14[0], r2
    vmov.32     d14[1], r3             @ multipliers
    mov         r12, r0
    movw        r0, #:lower16:ChanBuf
    movw        r2, #:lower16:SSumLR
    movt        r0, #:upper16:ChanBuf
    movt        r2, #:upper16:SSumLR
    add         r0, r12, lsl #2
    add         r2, r12, lsl #3
0:
    vldmia      r0!, {d0-d1}
    vldmia      r2, {d2-d5}
    vmul.s32    d10, d14, d0[0]
    vmul.s32    d11, d14, d0[1]
    vmul.s32    d12, d14, d1[0]
    vmul.s32    d13, d14, d1[1]
    vsra.s32    q1, q5, #14
    vsra.s32    q2, q6, #14
    subs        r1, #4
    blt         mc_finish
    vstmia      r2!, {d2-d5}
    bgt         0b
    nop
    bxeq        lr

mc_finish:
    vstmia      r2!, {d2}
    cmp         r1, #-2
    vstmiage    r2!, {d3}
    cmp         r1, #-1
    vstmiage    r2!, {d4}
    bx          lr


.global mix_chan_rvb @ (int start, int count, int lv, int rv)
mix_chan_rvb:
    vmov.32     d14[0], r2
    vmov.32     d14[1], r3             @ multipliers
    mov         r12, r0
    movw        r0, #:lower16:ChanBuf
    movw        r3, #:lower16:sRVBStart
    movw        r2, #:lower16:SSumLR
    movt        r0, #:upper16:ChanBuf
    movt        r3, #:upper16:sRVBStart
    movt        r2, #:upper16:SSumLR
    ldr         r3, [r3]
    add         r0, r12, lsl #2
    add         r2, r12, lsl #3
    add         r3, r12, lsl #3
0:
    vldmia      r0!, {d0-d1}
    vldmia      r2, {d2-d5}
    vldmia      r3, {d6-d9}
    vmul.s32    d10, d14, d0[0]
    vmul.s32    d11, d14, d0[1]
    vmul.s32    d12, d14, d1[0]
    vmul.s32    d13, d14, d1[1]
    vsra.s32    q1, q5, #14
    vsra.s32    q2, q6, #14
    vsra.s32    q3, q5, #14
    vsra.s32    q4, q6, #14
    subs        r1, #4
    blt         mcr_finish
    vstmia      r2!, {d2-d5}
    vstmia      r3!, {d6-d9}
    bgt         0b
    nop
    bxeq        lr

mcr_finish:
    vstmia      r2!, {d2}
    vstmia      r3!, {d6}
    cmp         r1, #-2
    vstmiage    r2!, {d3}
    vstmiage    r3!, {d7}
    cmp         r1, #-1
    vstmiage    r2!, {d4}
    vstmiage    r3!, {d8}
    bx          lr

@ vim:filetype=armasm
Commit	Line	Data
b17618c0	1	/*
	2	* (C) Gražvydas "notaz" Ignotas, 2011
	3	*
	4	* This work is licensed under the terms of any of these licenses
	5	* (at your option):
	6	* - GNU GPL, version 2 or later.
	7	* - GNU LGPL, version 2.1 or later.
	8	* See the COPYING file in the top-level directory.
	9	*/
	10
	11
	12	.text
	13	.align 2
	14
	15
	16	.global mix_chan @ (int start, int count, int lv, int rv)
	17	mix_chan:
	18	vmov.32 d14[0], r2
	19	vmov.32 d14[1], r3 @ multipliers
	20	mov r12, r0
	21	movw r0, #:lower16:ChanBuf
	22	movw r2, #:lower16:SSumLR
	23	movt r0, #:upper16:ChanBuf
	24	movt r2, #:upper16:SSumLR
	25	add r0, r12, lsl #2
	26	add r2, r12, lsl #3
	27	0:
	28	vldmia r0!, {d0-d1}
	29	vldmia r2, {d2-d5}
	30	vmul.s32 d10, d14, d0[0]
	31	vmul.s32 d11, d14, d0[1]
	32	vmul.s32 d12, d14, d1[0]
	33	vmul.s32 d13, d14, d1[1]
	34	vsra.s32 q1, q5, #14
	35	vsra.s32 q2, q6, #14
	36	subs r1, #4
	37	blt mc_finish
	38	vstmia r2!, {d2-d5}
	39	bgt 0b
	40	nop
	41	bxeq lr
	42
	43	mc_finish:
	44	vstmia r2!, {d2}
587fa7de	45	cmp r1, #-2
b17618c0	46	vstmiage r2!, {d3}
587fa7de	47	cmp r1, #-1
b17618c0	48	vstmiage r2!, {d4}
	49	bx lr
	50
	51
	52	.global mix_chan_rvb @ (int start, int count, int lv, int rv)
	53	mix_chan_rvb:
	54	vmov.32 d14[0], r2
	55	vmov.32 d14[1], r3 @ multipliers
	56	mov r12, r0
	57	movw r0, #:lower16:ChanBuf
	58	movw r3, #:lower16:sRVBStart
	59	movw r2, #:lower16:SSumLR
	60	movt r0, #:upper16:ChanBuf
	61	movt r3, #:upper16:sRVBStart
	62	movt r2, #:upper16:SSumLR
	63	ldr r3, [r3]
	64	add r0, r12, lsl #2
	65	add r2, r12, lsl #3
	66	add r3, r12, lsl #3
	67	0:
	68	vldmia r0!, {d0-d1}
	69	vldmia r2, {d2-d5}
	70	vldmia r3, {d6-d9}
	71	vmul.s32 d10, d14, d0[0]
	72	vmul.s32 d11, d14, d0[1]
	73	vmul.s32 d12, d14, d1[0]
	74	vmul.s32 d13, d14, d1[1]
	75	vsra.s32 q1, q5, #14
	76	vsra.s32 q2, q6, #14
	77	vsra.s32 q3, q5, #14
	78	vsra.s32 q4, q6, #14
	79	subs r1, #4
	80	blt mcr_finish
	81	vstmia r2!, {d2-d5}
	82	vstmia r3!, {d6-d9}
	83	bgt 0b
	84	nop
	85	bxeq lr
	86
	87	mcr_finish:
	88	vstmia r2!, {d2}
	89	vstmia r3!, {d6}
587fa7de	90	cmp r1, #-2
b17618c0	91	vstmiage r2!, {d3}
b17618c0	92	vstmiage r3!, {d7}
587fa7de	93	cmp r1, #-1
b17618c0	94	vstmiage r2!, {d4}
	95	vstmiage r3!, {d8}
	96	bx lr
	97
	98	@ vim:filetype=armasm