WebSVN – Kolibri OS – Blame – /contrib/sdk/sources/ffmpeg/ffmpeg-2.8/libswresample/x86/audio_convert.asm

Rev	Author	Line No.	Line
6147	serge	1	;******************************************************************************
		2	;* Copyright (c) 2012 Michael Niedermayer
		3	;*
		4	;* This file is part of FFmpeg.
		5	;*
		6	;* FFmpeg is free software; you can redistribute it and/or
		7	;* modify it under the terms of the GNU Lesser General Public
		8	;* License as published by the Free Software Foundation; either
		9	;* version 2.1 of the License, or (at your option) any later version.
		10	;*
		11	;* FFmpeg is distributed in the hope that it will be useful,
		12	;* but WITHOUT ANY WARRANTY; without even the implied warranty of
		13	;* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
		14	;* Lesser General Public License for more details.
		15	;*
		16	;* You should have received a copy of the GNU Lesser General Public
		17	;* License along with FFmpeg; if not, write to the Free Software
		18	;* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
		19	;******************************************************************************
		20
		21	%include "libavutil/x86/x86util.asm"
		22
		23	SECTION_RODATA 32
		24	flt2pm31: times 8 dd 4.6566129e-10
		25	flt2p31 : times 8 dd 2147483648.0
		26	flt2p15 : times 8 dd 32768.0
		27
		28	word_unpack_shuf : db 0, 1, 4, 5, 8, 9,12,13, 2, 3, 6, 7,10,11,14,15
		29
		30	SECTION .text
		31
		32
		33	;to, from, a/u, log2_outsize, log_intsize, const
		34	%macro PACK_2CH 5-7
		35	cglobal pack_2ch_%2_to_%1_%3, 3, 4, 6, dst, src, len, src2
		36	mov src2q , [srcq+gprsize]
		37	mov srcq , [srcq]
		38	mov dstq , [dstq]
		39	%ifidn %3, a
		40	test dstq, mmsize-1
		41	jne pack_2ch_%2_to_%1_u_int %+ SUFFIX
		42	test srcq, mmsize-1
		43	jne pack_2ch_%2_to_%1_u_int %+ SUFFIX
		44	test src2q, mmsize-1
		45	jne pack_2ch_%2_to_%1_u_int %+ SUFFIX
		46	%else
		47	pack_2ch_%2_to_%1_u_int %+ SUFFIX:
		48	%endif
		49	lea srcq , [srcq + (1<<%5)*lenq]
		50	lea src2q, [src2q + (1<<%5)*lenq]
		51	lea dstq , [dstq + (2<<%4)*lenq]
		52	neg lenq
		53	%7 m0,m1,m2,m3,m4,m5
		54	.next:
		55	%if %4 >= %5
		56	mov%3 m0, [ srcq +(1<<%5)*lenq]
		57	mova m1, m0
		58	mov%3 m2, [ src2q+(1<<%5)*lenq]
		59	%if %5 == 1
		60	punpcklwd m0, m2
		61	punpckhwd m1, m2
		62	%else
		63	punpckldq m0, m2
		64	punpckhdq m1, m2
		65	%endif
		66	%6 m0,m1,m2,m3,m4,m5
		67	%else
		68	mov%3 m0, [ srcq +(1<<%5)*lenq]
		69	mov%3 m1, [mmsize + srcq +(1<<%5)*lenq]
		70	mov%3 m2, [ src2q+(1<<%5)*lenq]
		71	mov%3 m3, [mmsize + src2q+(1<<%5)*lenq]
		72	%6 m0,m1,m2,m3,m4,m5
		73	mova m2, m0
		74	punpcklwd m0, m1
		75	punpckhwd m2, m1
		76	SWAP 1,2
		77	%endif
		78	mov%3 [ dstq+(2<<%4)*lenq], m0
		79	mov%3 [ mmsize + dstq+(2<<%4)*lenq], m1
		80	%if %4 > %5
		81	mov%3 [2mmsize + dstq+(2<<%4)lenq], m2
		82	mov%3 [3mmsize + dstq+(2<<%4)lenq], m3
		83	add lenq, 4*mmsize/(2<<%4)
		84	%else
		85	add lenq, 2*mmsize/(2<<%4)
		86	%endif
		87	jl .next
		88	REP_RET
		89	%endmacro
		90
		91	%macro UNPACK_2CH 5-7
		92	cglobal unpack_2ch_%2_to_%1_%3, 3, 4, 7, dst, src, len, dst2
		93	mov dst2q , [dstq+gprsize]
		94	mov srcq , [srcq]
		95	mov dstq , [dstq]
		96	%ifidn %3, a
		97	test dstq, mmsize-1
		98	jne unpack_2ch_%2_to_%1_u_int %+ SUFFIX
		99	test srcq, mmsize-1
		100	jne unpack_2ch_%2_to_%1_u_int %+ SUFFIX
		101	test dst2q, mmsize-1
		102	jne unpack_2ch_%2_to_%1_u_int %+ SUFFIX
		103	%else
		104	unpack_2ch_%2_to_%1_u_int %+ SUFFIX:
		105	%endif
		106	lea srcq , [srcq + (2<<%5)*lenq]
		107	lea dstq , [dstq + (1<<%4)*lenq]
		108	lea dst2q, [dst2q + (1<<%4)*lenq]
		109	neg lenq
		110	%7 m0,m1,m2,m3,m4,m5
		111	mova m6, [word_unpack_shuf]
		112	.next:
		113	mov%3 m0, [ srcq +(2<<%5)*lenq]
		114	mov%3 m2, [ mmsize + srcq +(2<<%5)*lenq]
		115	%if %5 == 1
		116	%ifidn SUFFIX, _ssse3
		117	pshufb m0, m6
		118	mova m1, m0
		119	pshufb m2, m6
		120	punpcklqdq m0,m2
		121	punpckhqdq m1,m2
		122	%else
		123	mova m1, m0
		124	punpcklwd m0,m2
		125	punpckhwd m1,m2
		126
		127	mova m2, m0
		128	punpcklwd m0,m1
		129	punpckhwd m2,m1
		130
		131	mova m1, m0
		132	punpcklwd m0,m2
		133	punpckhwd m1,m2
		134	%endif
		135	%else
		136	mova m1, m0
		137	shufps m0, m2, 10001000b
		138	shufps m1, m2, 11011101b
		139	%endif
		140	%if %4 < %5
		141	mov%3 m2, [2mmsize + srcq +(2<<%5)lenq]
		142	mova m3, m2
		143	mov%3 m4, [3mmsize + srcq +(2<<%5)lenq]
		144	shufps m2, m4, 10001000b
		145	shufps m3, m4, 11011101b
		146	SWAP 1,2
		147	%endif
		148	%6 m0,m1,m2,m3,m4,m5
		149	mov%3 [ dstq+(1<<%4)*lenq], m0
		150	%if %4 > %5
		151	mov%3 [ dst2q+(1<<%4)*lenq], m2
		152	mov%3 [ mmsize + dstq+(1<<%4)*lenq], m1
		153	mov%3 [ mmsize + dst2q+(1<<%4)*lenq], m3
		154	add lenq, 2*mmsize/(1<<%4)
		155	%else
		156	mov%3 [ dst2q+(1<<%4)*lenq], m1
		157	add lenq, mmsize/(1<<%4)
		158	%endif
		159	jl .next
		160	REP_RET
		161	%endmacro
		162
		163	%macro CONV 5-7
		164	cglobal %2_to_%1_%3, 3, 3, 6, dst, src, len
		165	mov srcq , [srcq]
		166	mov dstq , [dstq]
		167	%ifidn %3, a
		168	test dstq, mmsize-1
		169	jne %2_to_%1_u_int %+ SUFFIX
		170	test srcq, mmsize-1
		171	jne %2_to_%1_u_int %+ SUFFIX
		172	%else
		173	%2_to_%1_u_int %+ SUFFIX:
		174	%endif
		175	lea srcq , [srcq + (1<<%5)*lenq]
		176	lea dstq , [dstq + (1<<%4)*lenq]
		177	neg lenq
		178	%7 m0,m1,m2,m3,m4,m5
		179	.next:
		180	mov%3 m0, [ srcq +(1<<%5)*lenq]
		181	mov%3 m1, [ mmsize + srcq +(1<<%5)*lenq]
		182	%if %4 < %5
		183	mov%3 m2, [2mmsize + srcq +(1<<%5)lenq]
		184	mov%3 m3, [3mmsize + srcq +(1<<%5)lenq]
		185	%endif
		186	%6 m0,m1,m2,m3,m4,m5
		187	mov%3 [ dstq+(1<<%4)*lenq], m0
		188	mov%3 [ mmsize + dstq+(1<<%4)*lenq], m1
		189	%if %4 > %5
		190	mov%3 [2mmsize + dstq+(1<<%4)lenq], m2
		191	mov%3 [3mmsize + dstq+(1<<%4)lenq], m3
		192	add lenq, 4*mmsize/(1<<%4)
		193	%else
		194	add lenq, 2*mmsize/(1<<%4)
		195	%endif
		196	jl .next
		197	%if mmsize == 8
		198	emms
		199	RET
		200	%else
		201	REP_RET
		202	%endif
		203	%endmacro
		204
		205	%macro PACK_6CH 5-7
		206	cglobal pack_6ch_%2_to_%1_%3, 2,8,7, dst, src, src1, src2, src3, src4, src5, len
		207	%if ARCH_X86_64
		208	mov lend, r2d
		209	%else
		210	%define lend dword r2m
		211	%endif
		212	mov src1q, [srcq+1*gprsize]
		213	mov src2q, [srcq+2*gprsize]
		214	mov src3q, [srcq+3*gprsize]
		215	mov src4q, [srcq+4*gprsize]
		216	mov src5q, [srcq+5*gprsize]
		217	mov srcq, [srcq]
		218	mov dstq, [dstq]
		219	%ifidn %3, a
		220	test dstq, mmsize-1
		221	jne pack_6ch_%2_to_%1_u_int %+ SUFFIX
		222	test srcq, mmsize-1
		223	jne pack_6ch_%2_to_%1_u_int %+ SUFFIX
		224	test src1q, mmsize-1
		225	jne pack_6ch_%2_to_%1_u_int %+ SUFFIX
		226	test src2q, mmsize-1
		227	jne pack_6ch_%2_to_%1_u_int %+ SUFFIX
		228	test src3q, mmsize-1
		229	jne pack_6ch_%2_to_%1_u_int %+ SUFFIX
		230	test src4q, mmsize-1
		231	jne pack_6ch_%2_to_%1_u_int %+ SUFFIX
		232	test src5q, mmsize-1
		233	jne pack_6ch_%2_to_%1_u_int %+ SUFFIX
		234	%else
		235	pack_6ch_%2_to_%1_u_int %+ SUFFIX:
		236	%endif
		237	sub src1q, srcq
		238	sub src2q, srcq
		239	sub src3q, srcq
		240	sub src4q, srcq
		241	sub src5q, srcq
		242	%7 x,x,x,x,m7,x
		243	.loop:
		244	mov%3 m0, [srcq ]
		245	mov%3 m1, [srcq+src1q]
		246	mov%3 m2, [srcq+src2q]
		247	mov%3 m3, [srcq+src3q]
		248	mov%3 m4, [srcq+src4q]
		249	mov%3 m5, [srcq+src5q]
		250	%if cpuflag(sse)
		251	SBUTTERFLYPS 0, 1, 6
		252	SBUTTERFLYPS 2, 3, 6
		253	SBUTTERFLYPS 4, 5, 6
		254
		255	%if cpuflag(avx)
		256	blendps m6, m4, m0, 1100b
		257	%else
		258	movaps m6, m4
		259	shufps m4, m0, q3210
		260	SWAP 4,6
		261	%endif
		262	movlhps m0, m2
		263	movhlps m4, m2
		264	%if cpuflag(avx)
		265	blendps m2, m5, m1, 1100b
		266	%else
		267	movaps m2, m5
		268	shufps m5, m1, q3210
		269	SWAP 2,5
		270	%endif
		271	movlhps m1, m3
		272	movhlps m5, m3
		273
		274	%6 m0,m6,x,x,m7,m3
		275	%6 m4,m1,x,x,m7,m3
		276	%6 m2,m5,x,x,m7,m3
		277
		278	mov %+ %3 %+ ps [dstq ], m0
		279	mov %+ %3 %+ ps [dstq+16], m6
		280	mov %+ %3 %+ ps [dstq+32], m4
		281	mov %+ %3 %+ ps [dstq+48], m1
		282	mov %+ %3 %+ ps [dstq+64], m2
		283	mov %+ %3 %+ ps [dstq+80], m5
		284	%else ; mmx
		285	SBUTTERFLY dq, 0, 1, 6
		286	SBUTTERFLY dq, 2, 3, 6
		287	SBUTTERFLY dq, 4, 5, 6
		288
		289	movq [dstq ], m0
		290	movq [dstq+ 8], m2
		291	movq [dstq+16], m4
		292	movq [dstq+24], m1
		293	movq [dstq+32], m3
		294	movq [dstq+40], m5
		295	%endif
		296	add srcq, mmsize
		297	add dstq, mmsize*6
		298	sub lend, mmsize/4
		299	jg .loop
		300	%if mmsize == 8
		301	emms
		302	RET
		303	%else
		304	REP_RET
		305	%endif
		306	%endmacro
		307
		308	%macro UNPACK_6CH 5-7
		309	cglobal unpack_6ch_%2_to_%1_%3, 2, 8, 8, dst, src, dst1, dst2, dst3, dst4, dst5, len
		310	%if ARCH_X86_64
		311	mov lend, r2d
		312	%else
		313	%define lend dword r2m
		314	%endif
		315	mov dst1q, [dstq+1*gprsize]
		316	mov dst2q, [dstq+2*gprsize]
		317	mov dst3q, [dstq+3*gprsize]
		318	mov dst4q, [dstq+4*gprsize]
		319	mov dst5q, [dstq+5*gprsize]
		320	mov dstq, [dstq]
		321	mov srcq, [srcq]
		322	%ifidn %3, a
		323	test dstq, mmsize-1
		324	jne unpack_6ch_%2_to_%1_u_int %+ SUFFIX
		325	test srcq, mmsize-1
		326	jne unpack_6ch_%2_to_%1_u_int %+ SUFFIX
		327	test dst1q, mmsize-1
		328	jne unpack_6ch_%2_to_%1_u_int %+ SUFFIX
		329	test dst2q, mmsize-1
		330	jne unpack_6ch_%2_to_%1_u_int %+ SUFFIX
		331	test dst3q, mmsize-1
		332	jne unpack_6ch_%2_to_%1_u_int %+ SUFFIX
		333	test dst4q, mmsize-1
		334	jne unpack_6ch_%2_to_%1_u_int %+ SUFFIX
		335	test dst5q, mmsize-1
		336	jne unpack_6ch_%2_to_%1_u_int %+ SUFFIX
		337	%else
		338	unpack_6ch_%2_to_%1_u_int %+ SUFFIX:
		339	%endif
		340	sub dst1q, dstq
		341	sub dst2q, dstq
		342	sub dst3q, dstq
		343	sub dst4q, dstq
		344	sub dst5q, dstq
		345	%7 x,x,x,x,m7,x
		346	.loop:
		347	mov%3 m0, [srcq ]
		348	mov%3 m1, [srcq+16]
		349	mov%3 m2, [srcq+32]
		350	mov%3 m3, [srcq+48]
		351	mov%3 m4, [srcq+64]
		352	mov%3 m5, [srcq+80]
		353
		354	SBUTTERFLYPS 0, 3, 6
		355	SBUTTERFLYPS 1, 4, 6
		356	SBUTTERFLYPS 2, 5, 6
		357	SBUTTERFLYPS 0, 4, 6
		358	SBUTTERFLYPS 3, 2, 6
		359	SBUTTERFLYPS 1, 5, 6
		360	SWAP 1, 4
		361	SWAP 2, 3
		362
		363	%6 m0,m1,x,x,m7,m6
		364	%6 m2,m3,x,x,m7,m6
		365	%6 m4,m5,x,x,m7,m6
		366
		367	mov %+ %3 %+ ps [dstq ], m0
		368	mov %+ %3 %+ ps [dstq+dst1q], m1
		369	mov %+ %3 %+ ps [dstq+dst2q], m2
		370	mov %+ %3 %+ ps [dstq+dst3q], m3
		371	mov %+ %3 %+ ps [dstq+dst4q], m4
		372	mov %+ %3 %+ ps [dstq+dst5q], m5
		373
		374	add srcq, mmsize*6
		375	add dstq, mmsize
		376	sub lend, mmsize/4
		377	jg .loop
		378	REP_RET
		379	%endmacro
		380
		381	%define PACK_8CH_GPRS (10 * ARCH_X86_64) + ((6 + HAVE_ALIGNED_STACK) * ARCH_X86_32)
		382
		383	%macro PACK_8CH 5-7
		384	cglobal pack_8ch_%2_to_%1_%3, 2,PACK_8CH_GPRS,10, ARCH_X86_32*48, dst, src, len, src1, src2, src3, src4, src5, src6, src7
		385	mov dstq, [dstq]
		386	%if ARCH_X86_32
		387	DEFINE_ARGS dst, src, src2, src3, src4, src5, src6
		388	%define lend dword r2m
		389	%define src1q r0q
		390	%define src1m dword [rsp+32]
		391	%if HAVE_ALIGNED_STACK == 0
		392	DEFINE_ARGS dst, src, src2, src3, src5, src6
		393	%define src4q r0q
		394	%define src4m dword [rsp+36]
		395	%endif
		396	%define src7q r0q
		397	%define src7m dword [rsp+40]
		398	mov dstm, dstq
		399	%endif
		400	mov src7q, [srcq+7*gprsize]
		401	mov src6q, [srcq+6*gprsize]
		402	%if ARCH_X86_32
		403	mov src7m, src7q
		404	%endif
		405	mov src5q, [srcq+5*gprsize]
		406	mov src4q, [srcq+4*gprsize]
		407	mov src3q, [srcq+3*gprsize]
		408	%if ARCH_X86_32 && HAVE_ALIGNED_STACK == 0
		409	mov src4m, src4q
		410	%endif
		411	mov src2q, [srcq+2*gprsize]
		412	mov src1q, [srcq+1*gprsize]
		413	mov srcq, [srcq]
		414	%ifidn %3, a
		415	%if ARCH_X86_32
		416	test dstmp, mmsize-1
		417	%else
		418	test dstq, mmsize-1
		419	%endif
		420	jne pack_8ch_%2_to_%1_u_int %+ SUFFIX
		421	test srcq, mmsize-1
		422	jne pack_8ch_%2_to_%1_u_int %+ SUFFIX
		423	test src1q, mmsize-1
		424	jne pack_8ch_%2_to_%1_u_int %+ SUFFIX
		425	test src2q, mmsize-1
		426	jne pack_8ch_%2_to_%1_u_int %+ SUFFIX
		427	test src3q, mmsize-1
		428	jne pack_8ch_%2_to_%1_u_int %+ SUFFIX
		429	%if ARCH_X86_32 && HAVE_ALIGNED_STACK == 0
		430	test src4m, mmsize-1
		431	%else
		432	test src4q, mmsize-1
		433	%endif
		434	jne pack_8ch_%2_to_%1_u_int %+ SUFFIX
		435	test src5q, mmsize-1
		436	jne pack_8ch_%2_to_%1_u_int %+ SUFFIX
		437	test src6q, mmsize-1
		438	jne pack_8ch_%2_to_%1_u_int %+ SUFFIX
		439	%if ARCH_X86_32
		440	test src7m, mmsize-1
		441	%else
		442	test src7q, mmsize-1
		443	%endif
		444	jne pack_8ch_%2_to_%1_u_int %+ SUFFIX
		445	%else
		446	pack_8ch_%2_to_%1_u_int %+ SUFFIX:
		447	%endif
		448	sub src1q, srcq
		449	sub src2q, srcq
		450	sub src3q, srcq
		451	%if ARCH_X86_64 \|\| HAVE_ALIGNED_STACK
		452	sub src4q, srcq
		453	%else
		454	sub src4m, srcq
		455	%endif
		456	sub src5q, srcq
		457	sub src6q, srcq
		458	%if ARCH_X86_64
		459	sub src7q, srcq
		460	%else
		461	mov src1m, src1q
		462	sub src7m, srcq
		463	%endif
		464
		465	%if ARCH_X86_64
		466	%7 x,x,x,x,m9,x
		467	%elifidn %1, int32
		468	%define m9 [flt2p31]
		469	%else
		470	%define m9 [flt2pm31]
		471	%endif
		472
		473	.loop:
		474	mov%3 m0, [srcq ]
		475	mov%3 m1, [srcq+src1q]
		476	mov%3 m2, [srcq+src2q]
		477	%if ARCH_X86_32 && HAVE_ALIGNED_STACK == 0
		478	mov src4q, src4m
		479	%endif
		480	mov%3 m3, [srcq+src3q]
		481	mov%3 m4, [srcq+src4q]
		482	mov%3 m5, [srcq+src5q]
		483	%if ARCH_X86_32
		484	mov src7q, src7m
		485	%endif
		486	mov%3 m6, [srcq+src6q]
		487	mov%3 m7, [srcq+src7q]
		488
		489	%if ARCH_X86_64
		490	TRANSPOSE8x4D 0, 1, 2, 3, 4, 5, 6, 7, 8
		491
		492	%6 m0,m1,x,x,m9,m8
		493	%6 m2,m3,x,x,m9,m8
		494	%6 m4,m5,x,x,m9,m8
		495	%6 m6,m7,x,x,m9,m8
		496
		497	mov%3 [dstq], m0
		498	%else
		499	mov dstq, dstm
		500
		501	TRANSPOSE8x4D 0, 1, 2, 3, 4, 5, 6, 7, [rsp], [rsp+16], 1
		502
		503	%6 m0,m1,x,x,m9,m2
		504	mova m2, [rsp]
		505	mov%3 [dstq], m0
		506	%6 m2,m3,x,x,m9,m0
		507	%6 m4,m5,x,x,m9,m0
		508	%6 m6,m7,x,x,m9,m0
		509
		510	%endif
		511
		512	mov%3 [dstq+16], m1
		513	mov%3 [dstq+32], m2
		514	mov%3 [dstq+48], m3
		515	mov%3 [dstq+64], m4
		516	mov%3 [dstq+80], m5
		517	mov%3 [dstq+96], m6
		518	mov%3 [dstq+112], m7
		519
		520	add srcq, mmsize
		521	add dstq, mmsize*8
		522	%if ARCH_X86_32
		523	mov dstm, dstq
		524	mov src1q, src1m
		525	%endif
		526	sub lend, mmsize/4
		527	jg .loop
		528	REP_RET
		529	%endmacro
		530
		531	%macro INT16_TO_INT32_N 6
		532	pxor m2, m2
		533	pxor m3, m3
		534	punpcklwd m2, m1
		535	punpckhwd m3, m1
		536	SWAP 4,0
		537	pxor m0, m0
		538	pxor m1, m1
		539	punpcklwd m0, m4
		540	punpckhwd m1, m4
		541	%endmacro
		542
		543	%macro INT32_TO_INT16_N 6
		544	psrad m0, 16
		545	psrad m1, 16
		546	psrad m2, 16
		547	psrad m3, 16
		548	packssdw m0, m1
		549	packssdw m2, m3
		550	SWAP 1,2
		551	%endmacro
		552
		553	%macro INT32_TO_FLOAT_INIT 6
		554	mova %5, [flt2pm31]
		555	%endmacro
		556	%macro INT32_TO_FLOAT_N 6
		557	cvtdq2ps %1, %1
		558	cvtdq2ps %2, %2
		559	mulps %1, %1, %5
		560	mulps %2, %2, %5
		561	%endmacro
		562
		563	%macro FLOAT_TO_INT32_INIT 6
		564	mova %5, [flt2p31]
		565	%endmacro
		566	%macro FLOAT_TO_INT32_N 6
		567	mulps %1, %5
		568	mulps %2, %5
		569	cvtps2dq %6, %1
		570	cmpps %1, %1, %5, 5
		571	paddd %1, %6
		572	cvtps2dq %6, %2
		573	cmpps %2, %2, %5, 5
		574	paddd %2, %6
		575	%endmacro
		576
		577	%macro INT16_TO_FLOAT_INIT 6
		578	mova m5, [flt2pm31]
		579	%endmacro
		580	%macro INT16_TO_FLOAT_N 6
		581	INT16_TO_INT32_N %1,%2,%3,%4,%5,%6
		582	cvtdq2ps m0, m0
		583	cvtdq2ps m1, m1
		584	cvtdq2ps m2, m2
		585	cvtdq2ps m3, m3
		586	mulps m0, m0, m5
		587	mulps m1, m1, m5
		588	mulps m2, m2, m5
		589	mulps m3, m3, m5
		590	%endmacro
		591
		592	%macro FLOAT_TO_INT16_INIT 6
		593	mova m5, [flt2p15]
		594	%endmacro
		595	%macro FLOAT_TO_INT16_N 6
		596	mulps m0, m5
		597	mulps m1, m5
		598	mulps m2, m5
		599	mulps m3, m5
		600	cvtps2dq m0, m0
		601	cvtps2dq m1, m1
		602	packssdw m0, m1
		603	cvtps2dq m1, m2
		604	cvtps2dq m3, m3
		605	packssdw m1, m3
		606	%endmacro
		607
		608	%macro NOP_N 0-6
		609	%endmacro
		610
		611	INIT_MMX mmx
		612	CONV int32, int16, u, 2, 1, INT16_TO_INT32_N, NOP_N
		613	CONV int32, int16, a, 2, 1, INT16_TO_INT32_N, NOP_N
		614	CONV int16, int32, u, 1, 2, INT32_TO_INT16_N, NOP_N
		615	CONV int16, int32, a, 1, 2, INT32_TO_INT16_N, NOP_N
		616
		617	PACK_6CH float, float, u, 2, 2, NOP_N, NOP_N
		618	PACK_6CH float, float, a, 2, 2, NOP_N, NOP_N
		619
		620	INIT_XMM sse
		621	PACK_6CH float, float, u, 2, 2, NOP_N, NOP_N
		622	PACK_6CH float, float, a, 2, 2, NOP_N, NOP_N
		623
		624	UNPACK_6CH float, float, u, 2, 2, NOP_N, NOP_N
		625	UNPACK_6CH float, float, a, 2, 2, NOP_N, NOP_N
		626
		627	INIT_XMM sse2
		628	CONV int32, int16, u, 2, 1, INT16_TO_INT32_N, NOP_N
		629	CONV int32, int16, a, 2, 1, INT16_TO_INT32_N, NOP_N
		630	CONV int16, int32, u, 1, 2, INT32_TO_INT16_N, NOP_N
		631	CONV int16, int32, a, 1, 2, INT32_TO_INT16_N, NOP_N
		632
		633	PACK_2CH int16, int16, u, 1, 1, NOP_N, NOP_N
		634	PACK_2CH int16, int16, a, 1, 1, NOP_N, NOP_N
		635	PACK_2CH int32, int32, u, 2, 2, NOP_N, NOP_N
		636	PACK_2CH int32, int32, a, 2, 2, NOP_N, NOP_N
		637	PACK_2CH int32, int16, u, 2, 1, INT16_TO_INT32_N, NOP_N
		638	PACK_2CH int32, int16, a, 2, 1, INT16_TO_INT32_N, NOP_N
		639	PACK_2CH int16, int32, u, 1, 2, INT32_TO_INT16_N, NOP_N
		640	PACK_2CH int16, int32, a, 1, 2, INT32_TO_INT16_N, NOP_N
		641
		642	UNPACK_2CH int16, int16, u, 1, 1, NOP_N, NOP_N
		643	UNPACK_2CH int16, int16, a, 1, 1, NOP_N, NOP_N
		644	UNPACK_2CH int32, int32, u, 2, 2, NOP_N, NOP_N
		645	UNPACK_2CH int32, int32, a, 2, 2, NOP_N, NOP_N
		646	UNPACK_2CH int32, int16, u, 2, 1, INT16_TO_INT32_N, NOP_N
		647	UNPACK_2CH int32, int16, a, 2, 1, INT16_TO_INT32_N, NOP_N
		648	UNPACK_2CH int16, int32, u, 1, 2, INT32_TO_INT16_N, NOP_N
		649	UNPACK_2CH int16, int32, a, 1, 2, INT32_TO_INT16_N, NOP_N
		650
		651	CONV float, int32, u, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		652	CONV float, int32, a, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		653	CONV int32, float, u, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		654	CONV int32, float, a, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		655	CONV float, int16, u, 2, 1, INT16_TO_FLOAT_N, INT16_TO_FLOAT_INIT
		656	CONV float, int16, a, 2, 1, INT16_TO_FLOAT_N, INT16_TO_FLOAT_INIT
		657	CONV int16, float, u, 1, 2, FLOAT_TO_INT16_N, FLOAT_TO_INT16_INIT
		658	CONV int16, float, a, 1, 2, FLOAT_TO_INT16_N, FLOAT_TO_INT16_INIT
		659
		660	PACK_2CH float, int32, u, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		661	PACK_2CH float, int32, a, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		662	PACK_2CH int32, float, u, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		663	PACK_2CH int32, float, a, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		664	PACK_2CH float, int16, u, 2, 1, INT16_TO_FLOAT_N, INT16_TO_FLOAT_INIT
		665	PACK_2CH float, int16, a, 2, 1, INT16_TO_FLOAT_N, INT16_TO_FLOAT_INIT
		666	PACK_2CH int16, float, u, 1, 2, FLOAT_TO_INT16_N, FLOAT_TO_INT16_INIT
		667	PACK_2CH int16, float, a, 1, 2, FLOAT_TO_INT16_N, FLOAT_TO_INT16_INIT
		668
		669	UNPACK_2CH float, int32, u, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		670	UNPACK_2CH float, int32, a, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		671	UNPACK_2CH int32, float, u, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		672	UNPACK_2CH int32, float, a, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		673	UNPACK_2CH float, int16, u, 2, 1, INT16_TO_FLOAT_N, INT16_TO_FLOAT_INIT
		674	UNPACK_2CH float, int16, a, 2, 1, INT16_TO_FLOAT_N, INT16_TO_FLOAT_INIT
		675	UNPACK_2CH int16, float, u, 1, 2, FLOAT_TO_INT16_N, FLOAT_TO_INT16_INIT
		676	UNPACK_2CH int16, float, a, 1, 2, FLOAT_TO_INT16_N, FLOAT_TO_INT16_INIT
		677
		678	PACK_6CH float, int32, u, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		679	PACK_6CH float, int32, a, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		680	PACK_6CH int32, float, u, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		681	PACK_6CH int32, float, a, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		682
		683	UNPACK_6CH float, int32, u, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		684	UNPACK_6CH float, int32, a, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		685	UNPACK_6CH int32, float, u, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		686	UNPACK_6CH int32, float, a, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		687
		688	PACK_8CH float, float, u, 2, 2, NOP_N, NOP_N
		689	PACK_8CH float, float, a, 2, 2, NOP_N, NOP_N
		690
		691	PACK_8CH float, int32, u, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		692	PACK_8CH float, int32, a, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		693	PACK_8CH int32, float, u, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		694	PACK_8CH int32, float, a, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		695
		696	INIT_XMM ssse3
		697	UNPACK_2CH int16, int16, u, 1, 1, NOP_N, NOP_N
		698	UNPACK_2CH int16, int16, a, 1, 1, NOP_N, NOP_N
		699	UNPACK_2CH int32, int16, u, 2, 1, INT16_TO_INT32_N, NOP_N
		700	UNPACK_2CH int32, int16, a, 2, 1, INT16_TO_INT32_N, NOP_N
		701	UNPACK_2CH float, int16, u, 2, 1, INT16_TO_FLOAT_N, INT16_TO_FLOAT_INIT
		702	UNPACK_2CH float, int16, a, 2, 1, INT16_TO_FLOAT_N, INT16_TO_FLOAT_INIT
		703
		704	%if HAVE_AVX_EXTERNAL
		705	INIT_XMM avx
		706	PACK_6CH float, float, u, 2, 2, NOP_N, NOP_N
		707	PACK_6CH float, float, a, 2, 2, NOP_N, NOP_N
		708
		709	UNPACK_6CH float, float, u, 2, 2, NOP_N, NOP_N
		710	UNPACK_6CH float, float, a, 2, 2, NOP_N, NOP_N
		711
		712	PACK_6CH float, int32, u, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		713	PACK_6CH float, int32, a, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		714	PACK_6CH int32, float, u, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		715	PACK_6CH int32, float, a, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		716
		717	UNPACK_6CH float, int32, u, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		718	UNPACK_6CH float, int32, a, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		719	UNPACK_6CH int32, float, u, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		720	UNPACK_6CH int32, float, a, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		721
		722	PACK_8CH float, float, u, 2, 2, NOP_N, NOP_N
		723	PACK_8CH float, float, a, 2, 2, NOP_N, NOP_N
		724
		725	PACK_8CH float, int32, u, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		726	PACK_8CH float, int32, a, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		727	PACK_8CH int32, float, u, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		728	PACK_8CH int32, float, a, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		729
		730	INIT_YMM avx
		731	CONV float, int32, u, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		732	CONV float, int32, a, 2, 2, INT32_TO_FLOAT_N, INT32_TO_FLOAT_INIT
		733	%endif
		734
		735	%if HAVE_AVX2_EXTERNAL
		736	INIT_YMM avx2
		737	CONV int32, float, u, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		738	CONV int32, float, a, 2, 2, FLOAT_TO_INT32_N, FLOAT_TO_INT32_INIT
		739	%endif

Subversion Repositories Kolibri OS

(root)/contrib/sdk/sources/ffmpeg/ffmpeg-2.8/libswresample/x86/audio_convert.asm – Rev 6147