libavcodec/mips: Fix specification of instruction name

1.'xor,or,and' to 'pxor,por,pand'. In the case of operating FPR, gcc supports both of them, clang only supports the second type. 2.'dsrl,srl' to 'ssrld,ssrlw'. In the case of operating FPR, gcc supports both of them, clang only supports the second type.

Jin Bo committed 5 years ago

Commit ebedd261 parent e41255c16 files · +397 −397Browse files

Showing the changes to libavcodec (16 of 16 files). Show all

Unified · Split

16 files

libavcodec/mips/blockdsp_mmi.c +4 −4
libavcodec/mips/h264chroma_mmi.c +10 −10
libavcodec/mips/h264dsp_mmi.c +144 −144
libavcodec/mips/h264pred_mmi.c +9 −9
libavcodec/mips/h264qpel_mmi.c +13 −13
libavcodec/mips/hevcdsp_mmi.c +16 −16
libavcodec/mips/hpeldsp_mmi.c +13 −13
libavcodec/mips/idctdsp_mmi.c +1 −1
libavcodec/mips/mpegvideo_mmi.c +47 −47
libavcodec/mips/pixblockdsp_mmi.c +4 −4
libavcodec/mips/simple_idct_mmi.c +7 −7
libavcodec/mips/vc1dsp_mmi.c +17 −17
libavcodec/mips/vp3dsp_idct_mmi.c +66 −66
libavcodec/mips/vp8dsp_mmi.c +40 −40
libavcodec/mips/vp9_mc_mmi.c +5 −5
libavcodec/mips/wmv2dsp_mmi.c +1 −1

libavcodec/mips/blockdsp_mmi.c+4−4

@@ -76,8 +76,8 @@ void ff_clear_block_mmi(int16_t *block) · 75 unmodified lines
76	76	double ftmp[2];
77	77
78	78	__asm__ volatile (
79		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
80		− "xor %[ftmp1], %[ftmp1], %[ftmp1] \n\t"
	79	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	80	+ "pxor %[ftmp1], %[ftmp1], %[ftmp1] \n\t"
81	81	MMI_SQC1(%[ftmp0], %[ftmp1], %[block], 0x00)
82	82	MMI_SQC1(%[ftmp0], %[ftmp1], %[block], 0x10)
83	83	MMI_SQC1(%[ftmp0], %[ftmp1], %[block], 0x20)
@@ -97,8 +97,8 @@ void ff_clear_blocks_mmi(int16_t *block) · 13 unmodified lines
97	97	double ftmp[2];
98	98
99	99	__asm__ volatile (
100		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
101		− "xor %[ftmp1], %[ftmp1], %[ftmp1] \n\t"
	100	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	101	+ "pxor %[ftmp1], %[ftmp1], %[ftmp1] \n\t"
102	102	MMI_SQC1(%[ftmp0], %[ftmp1], %[block], 0x00)
103	103	MMI_SQC1(%[ftmp0], %[ftmp1], %[block], 0x10)
104	104	MMI_SQC1(%[ftmp0], %[ftmp1], %[block], 0x20)

libavcodec/mips/h264chroma_mmi.c+10−10

@@ -72,7 +72,7 @@ void ff_put_h264_chroma_mc8_mmi(uint8_t dst, uint8_t src, ptrdiff_t stride, · 71 unmodified lines
72	72	A = 64 - D - B - C;
73	73
74	74	__asm__ volatile (
75		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	75	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
76	76	"dli %[tmp0], 0x06 \n\t"
77	77	"pshufh %[A], %[A], %[ftmp0] \n\t"
78	78	"pshufh %[B], %[B], %[ftmp0] \n\t"
@@ -172,7 +172,7 @@ void ff_put_h264_chroma_mc8_mmi(uint8_t dst, uint8_t src, ptrdiff_t stride, · 93 unmodified lines
172	172	A = 64 - E;
173	173
174	174	__asm__ volatile (
175		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	175	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
176	176	"dli %[tmp0], 0x06 \n\t"
177	177	"pshufh %[A], %[A], %[ftmp0] \n\t"
178	178	"pshufh %[E], %[E], %[ftmp0] \n\t"
@@ -221,7 +221,7 @@ void ff_put_h264_chroma_mc8_mmi(uint8_t dst, uint8_t src, ptrdiff_t stride, · 42 unmodified lines
221	221	A = 64 - E;
222	222
223	223	__asm__ volatile (
224		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	224	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
225	225	"dli %[tmp0], 0x06 \n\t"
226	226	"pshufh %[A], %[A], %[ftmp0] \n\t"
227	227	"pshufh %[E], %[E], %[ftmp0] \n\t"
@@ -328,7 +328,7 @@ void ff_avg_h264_chroma_mc8_mmi(uint8_t dst, uint8_t src, ptrdiff_t stride, · 100 unmodified lines
328	328	C = (y << 3) - D;
329	329	A = 64 - D - B - C;
330	330	__asm__ volatile (
331		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	331	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
332	332	"dli %[tmp0], 0x06 \n\t"
333	333	"pshufh %[A], %[A], %[ftmp0] \n\t"
334	334	"pshufh %[B], %[B], %[ftmp0] \n\t"
@@ -396,7 +396,7 @@ void ff_avg_h264_chroma_mc8_mmi(uint8_t dst, uint8_t src, ptrdiff_t stride, · 61 unmodified lines
396	396	E = x << 3;
397	397	A = 64 - E;
398	398	__asm__ volatile (
399		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	399	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
400	400	"dli %[tmp0], 0x06 \n\t"
401	401	"pshufh %[A], %[A], %[ftmp0] \n\t"
402	402	"pshufh %[E], %[E], %[ftmp0] \n\t"
@@ -446,7 +446,7 @@ void ff_avg_h264_chroma_mc8_mmi(uint8_t dst, uint8_t src, ptrdiff_t stride, · 43 unmodified lines
446	446	E = y << 3;
447	447	A = 64 - E;
448	448	__asm__ volatile (
449		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	449	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
450	450	"dli %[tmp0], 0x06 \n\t"
451	451	"pshufh %[A], %[A], %[ftmp0] \n\t"
452	452	"pshufh %[E], %[E], %[ftmp0] \n\t"
@@ -509,7 +509,7 @@ void ff_put_h264_chroma_mc4_mmi(uint8_t dst, uint8_t src, ptrdiff_t stride, · 56 unmodified lines
509	509
510	510	if (D) {
511	511	__asm__ volatile (
512		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	512	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
513	513	"dli %[tmp0], 0x06 \n\t"
514	514	"pshufh %[A], %[A], %[ftmp0] \n\t"
515	515	"pshufh %[B], %[B], %[ftmp0] \n\t"
@@ -559,7 +559,7 @@ void ff_put_h264_chroma_mc4_mmi(uint8_t dst, uint8_t src, ptrdiff_t stride, · 43 unmodified lines
559	559	} else if (E) {
560	560	const int step = C ? stride : 1;
561	561	__asm__ volatile (
562		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	562	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
563	563	"dli %[tmp0], 0x06 \n\t"
564	564	"pshufh %[A], %[A], %[ftmp0] \n\t"
565	565	"pshufh %[E], %[E], %[ftmp0] \n\t"
@@ -633,7 +633,7 @@ void ff_avg_h264_chroma_mc4_mmi(uint8_t dst, uint8_t src, ptrdiff_t stride, · 67 unmodified lines
633	633
634	634	if (D) {
635	635	__asm__ volatile (
636		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	636	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
637	637	"dli %[tmp0], 0x06 \n\t"
638	638	"pshufh %[A], %[A], %[ftmp0] \n\t"
639	639	"pshufh %[B], %[B], %[ftmp0] \n\t"
@@ -685,7 +685,7 @@ void ff_avg_h264_chroma_mc4_mmi(uint8_t dst, uint8_t src, ptrdiff_t stride, · 45 unmodified lines
685	685	} else if (E) {
686	686	const int step = C ? stride : 1;
687	687	__asm__ volatile (
688		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	688	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
689	689	"dli %[tmp0], 0x06 \n\t"
690	690	"pshufh %[A], %[A], %[ftmp0] \n\t"
691	691	"pshufh %[E], %[E], %[ftmp0] \n\t"

libavcodec/mips/h264dsp_mmi.c+144−144

@@ -34,7 +34,7 @@ void ff_h264_add_pixels4_8_mmi(uint8_t dst, int16_t src, int stride) · 33 unmodified lines
34	34	DECLARE_VAR_LOW32;
35	35
36	36	__asm__ volatile (
37		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	37	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
38	38	MMI_LDC1(%[ftmp1], %[src], 0x00)
39	39	MMI_LDC1(%[ftmp2], %[src], 0x08)
40	40	MMI_LDC1(%[ftmp3], %[src], 0x10)
@@ -89,7 +89,7 @@ void ff_h264_idct_add_8_mmi(uint8_t dst, int16_t block, int stride) · 48 unmodified lines
89	89	MMI_LDC1(%[ftmp2], %[block], 0x10)
90	90	MMI_LDC1(%[ftmp3], %[block], 0x18)
91	91	/* memset(block, 0, 32) */
92		− "xor %[ftmp4], %[ftmp4], %[ftmp4] \n\t"
	92	+ "pxor %[ftmp4], %[ftmp4], %[ftmp4] \n\t"
93	93	"gssqc1 %[ftmp4], %[ftmp4], 0x00(%[block]) \n\t"
94	94	"gssqc1 %[ftmp4], %[ftmp4], 0x10(%[block]) \n\t"
95	95	"dli %[tmp0], 0x01 \n\t"
@@ -127,7 +127,7 @@ void ff_h264_idct_add_8_mmi(uint8_t dst, int16_t block, int stride) · 31 unmodified lines
127	127	"psubh %[ftmp5], %[ftmp5], %[ftmp4] \n\t"
128	128	MMI_ULWC1(%[ftmp2], %[dst], 0x00)
129	129	MMI_LWXC1(%[ftmp0], %[dst], %[stride], 0x00)
130		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
	130	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
131	131	"psrah %[ftmp3], %[ftmp10], %[ftmp9] \n\t"
132	132	"psrah %[ftmp4], %[ftmp11], %[ftmp9] \n\t"
133	133	"punpcklbh %[ftmp2], %[ftmp2], %[ftmp7] \n\t"
@@ -419,7 +419,7 @@ void ff_h264_idct8_add_8_mmi(uint8_t dst, int16_t block, int stride) · 285 unmodified lines
419	419	"psubh %[ftmp7], %[ftmp7], %[ftmp1] \n\t"
420	420	MMI_SDC1(%[ftmp0], $sp, 0x10)
421	421	"dmfc1 %[tmp1], %[ftmp2] \n\t"
422		− "xor %[ftmp2], %[ftmp2], %[ftmp2] \n\t"
	422	+ "pxor %[ftmp2], %[ftmp2], %[ftmp2] \n\t"
423	423	MMI_SDC1(%[ftmp2], %[block], 0x00)
424	424	MMI_SDC1(%[ftmp2], %[block], 0x08)
425	425	MMI_SDC1(%[ftmp2], %[block], 0x10)
@@ -555,7 +555,7 @@ void ff_h264_idct8_add_8_mmi(uint8_t dst, int16_t block, int stride) · 129 unmodified lines
555	555	"psubh %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
556	556	MMI_SDC1(%[ftmp7], $sp, 0x18)
557	557	"dmfc1 %[tmp2], %[ftmp0] \n\t"
558		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	558	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
559	559	MMI_ULWC1(%[ftmp6], %[addr0], 0x00)
560	560	MMI_LWXC1(%[ftmp7], %[addr0], %[stride], 0x00)
561	561	"psrah %[ftmp2], %[ftmp2], %[ftmp10] \n\t"
@@ -646,7 +646,7 @@ void ff_h264_idct_dc_add_8_mmi(uint8_t dst, int16_t block, int stride) · 84 unmodified lines
646	646
647	647	__asm__ volatile (
648	648	"mtc1 %[dc], %[ftmp5] \n\t"
649		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	649	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
650	650	"pshufh %[ftmp5], %[ftmp5], %[ftmp0] \n\t"
651	651	MMI_ULWC1(%[ftmp1], %[dst0], 0x00)
652	652	MMI_ULWC1(%[ftmp2], %[dst1], 0x00)
@@ -690,7 +690,7 @@ void ff_h264_idct8_dc_add_8_mmi(uint8_t dst, int16_t block, int stride) · 37 unmodified lines
690	690
691	691	__asm__ volatile (
692	692	"mtc1 %[dc], %[ftmp5] \n\t"
693		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	693	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
694	694	"pshufh %[ftmp5], %[ftmp5], %[ftmp0] \n\t"
695	695	MMI_LDC1(%[ftmp1], %[dst0], 0x00)
696	696	MMI_LDC1(%[ftmp2], %[dst1], 0x00)
@@ -929,7 +929,7 @@ void ff_h264_luma_dc_dequant_idct_8_mmi(int16_t output, int16_t input, · 232 unmodified lines
929	929	"packsswh %[ftmp0], %[ftmp0], %[ftmp1] \n\t"
930	930	"packsswh %[ftmp2], %[ftmp2], %[ftmp5] \n\t"
931	931	"dmfc1 %[tmp1], %[ftmp0] \n\t"
932		− "dsrl %[ftmp0], %[ftmp0], %[ftmp9] \n\t"
	932	+ "ssrld %[ftmp0], %[ftmp0], %[ftmp9] \n\t"
933	933	"mfc1 %[input], %[ftmp0] \n\t"
934	934	"sh %[tmp1], 0x00(%[output]) \n\t"
935	935	"sh %[input], 0x80(%[output]) \n\t"
@@ -938,7 +938,7 @@ void ff_h264_luma_dc_dequant_idct_8_mmi(int16_t output, int16_t input, · 2 unmodified lines
938	938	"sh %[tmp1], 0x20(%[output]) \n\t"
939	939	"sh %[input], 0xa0(%[output]) \n\t"
940	940	"dmfc1 %[tmp1], %[ftmp2] \n\t"
941		− "dsrl %[ftmp2], %[ftmp2], %[ftmp9] \n\t"
	941	+ "ssrld %[ftmp2], %[ftmp2], %[ftmp9] \n\t"
942	942	"mfc1 %[input], %[ftmp2] \n\t"
943	943	"sh %[tmp1], 0x40(%[output]) \n\t"
944	944	"sh %[input], 0xc0(%[output]) \n\t"
@@ -963,7 +963,7 @@ void ff_h264_luma_dc_dequant_idct_8_mmi(int16_t output, int16_t input, · 18 unmodified lines
963	963	"packsswh %[ftmp3], %[ftmp3], %[ftmp1] \n\t"
964	964	"packsswh %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
965	965	"dmfc1 %[tmp1], %[ftmp3] \n\t"
966		− "dsrl %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
	966	+ "ssrld %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
967	967	"mfc1 %[input], %[ftmp3] \n\t"
968	968	"sh %[tmp1], 0x100(%[output]) \n\t"
969	969	"sh %[input], 0x180(%[output]) \n\t"
@@ -972,7 +972,7 @@ void ff_h264_luma_dc_dequant_idct_8_mmi(int16_t output, int16_t input, · 2 unmodified lines
972	972	"sh %[tmp1], 0x120(%[output]) \n\t"
973	973	"sh %[input], 0x1a0(%[output]) \n\t"
974	974	"dmfc1 %[tmp1], %[ftmp4] \n\t"
975		− "dsrl %[ftmp4], %[ftmp4], %[ftmp9] \n\t"
	975	+ "ssrld %[ftmp4], %[ftmp4], %[ftmp9] \n\t"
976	976	"mfc1 %[input], %[ftmp4] \n\t"
977	977	"sh %[tmp1], 0x140(%[output]) \n\t"
978	978	"sh %[input], 0x1c0(%[output]) \n\t"
@@ -1016,7 +1016,7 @@ void ff_h264_luma_dc_dequant_idct_8_mmi(int16_t output, int16_t input, · 37 unmodified lines
1016	1016	"packsswh %[ftmp0], %[ftmp0], %[ftmp1] \n\t"
1017	1017	"packsswh %[ftmp2], %[ftmp2], %[ftmp5] \n\t"
1018	1018	"dmfc1 %[tmp1], %[ftmp0] \n\t"
1019		− "dsrl %[ftmp0], %[ftmp0], %[ftmp9] \n\t"
	1019	+ "ssrld %[ftmp0], %[ftmp0], %[ftmp9] \n\t"
1020	1020	"sh %[tmp1], 0x00(%[output]) \n\t"
1021	1021	"mfc1 %[input], %[ftmp0] \n\t"
1022	1022	"dsrl %[tmp1], %[tmp1], 0x10 \n\t"
@@ -1025,7 +1025,7 @@ void ff_h264_luma_dc_dequant_idct_8_mmi(int16_t output, int16_t input, · 2 unmodified lines
1025	1025	PTR_SRL "%[input], %[input], 0x10 \n\t"
1026	1026	"dmfc1 %[tmp1], %[ftmp2] \n\t"
1027	1027	"sh %[input], 0xa0(%[output]) \n\t"
1028		− "dsrl %[ftmp2], %[ftmp2], %[ftmp9] \n\t"
	1028	+ "ssrld %[ftmp2], %[ftmp2], %[ftmp9] \n\t"
1029	1029	"sh %[tmp1], 0x40(%[output]) \n\t"
1030	1030	"mfc1 %[input], %[ftmp2] \n\t"
1031	1031	"dsrl %[tmp1], %[tmp1], 0x10 \n\t"
@@ -1050,7 +1050,7 @@ void ff_h264_luma_dc_dequant_idct_8_mmi(int16_t output, int16_t input, · 18 unmodified lines
1050	1050	"packsswh %[ftmp3], %[ftmp3], %[ftmp1] \n\t"
1051	1051	"packsswh %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
1052	1052	"dmfc1 %[tmp1], %[ftmp3] \n\t"
1053		− "dsrl %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
	1053	+ "ssrld %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
1054	1054	"mfc1 %[input], %[ftmp3] \n\t"
1055	1055	"sh %[tmp1], 0x100(%[output]) \n\t"
1056	1056	"sh %[input], 0x180(%[output]) \n\t"
@@ -1059,7 +1059,7 @@ void ff_h264_luma_dc_dequant_idct_8_mmi(int16_t output, int16_t input, · 2 unmodified lines
1059	1059	"sh %[tmp1], 0x120(%[output]) \n\t"
1060	1060	"sh %[input], 0x1a0(%[output]) \n\t"
1061	1061	"dmfc1 %[tmp1], %[ftmp4] \n\t"
1062		− "dsrl %[ftmp4], %[ftmp4], %[ftmp9] \n\t"
	1062	+ "ssrld %[ftmp4], %[ftmp4], %[ftmp9] \n\t"
1063	1063	"mfc1 %[input], %[ftmp4] \n\t"
1064	1064	"sh %[tmp1], 0x140(%[output]) \n\t"
1065	1065	"sh %[input], 0x1c0(%[output]) \n\t"
@@ -1144,7 +1144,7 @@ void ff_h264_weight_pixels16_8_mmi(uint8_t *block, ptrdiff_t stride, int height, · 78 unmodified lines
1144	1144
1145	1145	for (y=0; y<height; y++, block+=stride) {
1146	1146	__asm__ volatile (
1147		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1147	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1148	1148	MMI_LDC1(%[ftmp1], %[block0], 0x00)
1149	1149	MMI_LDC1(%[ftmp2], %[block1], 0x00)
1150	1150	"mtc1 %[weight], %[ftmp3] \n\t"
@@ -1198,7 +1198,7 @@ void ff_h264_biweight_pixels16_8_mmi(uint8_t dst, uint8_t src, · 47 unmodified lines
1198	1198
1199	1199	for (y=0; y<height; y++, dst+=stride, src+=stride) {
1200	1200	__asm__ volatile (
1201		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1201	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1202	1202	MMI_LDC1(%[ftmp1], %[src0], 0x00)
1203	1203	MMI_LDC1(%[ftmp2], %[dst0], 0x00)
1204	1204	"mtc1 %[weights], %[ftmp3] \n\t"
@@ -1271,7 +1271,7 @@ void ff_h264_weight_pixels8_8_mmi(uint8_t *block, ptrdiff_t stride, int height, · 66 unmodified lines
1271	1271
1272	1272	for (y=0; y<height; y++, block+=stride) {
1273	1273	__asm__ volatile (
1274		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1274	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1275	1275	MMI_LDC1(%[ftmp1], %[block], 0x00)
1276	1276	"mtc1 %[weight], %[ftmp2] \n\t"
1277	1277	"mtc1 %[offset], %[ftmp3] \n\t"
@@ -1312,7 +1312,7 @@ void ff_h264_biweight_pixels8_8_mmi(uint8_t dst, uint8_t src, · 34 unmodified lines
1312	1312
1313	1313	for (y=0; y<height; y++, dst+=stride, src+=stride) {
1314	1314	__asm__ volatile (
1315		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1315	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1316	1316	MMI_LDC1(%[ftmp1], %[src], 0x00)
1317	1317	MMI_LDC1(%[ftmp2], %[dst], 0x00)
1318	1318	"mtc1 %[weights], %[ftmp3] \n\t"
@@ -1366,7 +1366,7 @@ void ff_h264_weight_pixels4_8_mmi(uint8_t *block, ptrdiff_t stride, int height, · 47 unmodified lines
1366	1366
1367	1367	for (y=0; y<height; y++, block+=stride) {
1368	1368	__asm__ volatile (
1369		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1369	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1370	1370	MMI_ULWC1(%[ftmp1], %[block], 0x00)
1371	1371	"mtc1 %[weight], %[ftmp2] \n\t"
1372	1372	"mtc1 %[offset], %[ftmp3] \n\t"
@@ -1402,7 +1402,7 @@ void ff_h264_biweight_pixels4_8_mmi(uint8_t dst, uint8_t src, · 29 unmodified lines
1402	1402
1403	1403	for (y=0; y<height; y++, dst+=stride, src+=stride) {
1404	1404	__asm__ volatile (
1405		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1405	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1406	1406	MMI_ULWC1(%[ftmp1], %[src], 0x00)
1407	1407	MMI_ULWC1(%[ftmp2], %[dst], 0x00)
1408	1408	"mtc1 %[weight], %[ftmp3] \n\t"
@@ -1445,7 +1445,7 @@ void ff_deblock_v8_luma_8_mmi(uint8_t *pix, ptrdiff_t stride, int alpha, int bet · 36 unmodified lines
1445	1445
1446	1446	__asm__ volatile (
1447	1447	PTR_ADDU "%[addr0], %[stride], %[stride] \n\t"
1448		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1448	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1449	1449	PTR_ADDU "%[addr1], %[stride], %[addr0] \n\t"
1450	1450	"addi %[alpha], %[alpha], -0x01 \n\t"
1451	1451	PTR_SUBU "%[addr1], $0, %[addr1] \n\t"
@@ -1463,18 +1463,18 @@ void ff_deblock_v8_luma_8_mmi(uint8_t *pix, ptrdiff_t stride, int alpha, int bet · 11 unmodified lines
1463	1463	"packushb %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
1464	1464	"psubusb %[ftmp7], %[ftmp3], %[ftmp2] \n\t"
1465	1465	"psubusb %[ftmp8], %[ftmp2], %[ftmp3] \n\t"
1466		− "or %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
	1466	+ "por %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
1467	1467	"psubusb %[ftmp7], %[ftmp2], %[ftmp1] \n\t"
1468	1468	"psubusb %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
1469	1469	"psubusb %[ftmp5], %[ftmp1], %[ftmp2] \n\t"
1470		− "or %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
	1470	+ "por %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1471	1471	"psubusb %[ftmp7], %[ftmp3], %[ftmp4] \n\t"
1472	1472	"psubusb %[ftmp5], %[ftmp5], %[ftmp6] \n\t"
1473		− "or %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
	1473	+ "por %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
1474	1474	"psubusb %[ftmp5], %[ftmp4], %[ftmp3] \n\t"
1475		− "or %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
	1475	+ "por %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1476	1476	"psubusb %[ftmp5], %[ftmp5], %[ftmp6] \n\t"
1477		− "or %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
	1477	+ "por %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
1478	1478	"pcmpeqb %[ftmp8], %[ftmp8], %[ftmp0] \n\t"
1479	1479	"pcmpeqb %[ftmp4], %[ftmp4], %[ftmp4] \n\t"
1480	1480	MMI_ULWC1(%[ftmp5], %[tc0], 0x00)
@@ -1482,21 +1482,21 @@ void ff_deblock_v8_luma_8_mmi(uint8_t *pix, ptrdiff_t stride, int alpha, int bet · 1 unmodified line
1482	1482	"punpcklbh %[ftmp9], %[ftmp5], %[ftmp5] \n\t"
1483	1483	"pcmpgtb %[ftmp5], %[ftmp9], %[ftmp4] \n\t"
1484	1484	MMI_LDC1(%[ftmp4], %[addr1], 0x00)
1485		− "and %[ftmp10], %[ftmp5], %[ftmp8] \n\t"
	1485	+ "pand %[ftmp10], %[ftmp5], %[ftmp8] \n\t"
1486	1486	"psubusb %[ftmp8], %[ftmp4], %[ftmp2] \n\t"
1487	1487	"psubusb %[ftmp7], %[ftmp2], %[ftmp4] \n\t"
1488	1488	"psubusb %[ftmp8], %[ftmp8], %[ftmp6] \n\t"
1489	1489	"psubusb %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
1490	1490	"pcmpeqb %[ftmp7], %[ftmp7], %[ftmp8] \n\t"
1491		− "and %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
1492		− "and %[ftmp5], %[ftmp10], %[ftmp9] \n\t"
	1491	+ "pand %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
	1492	+ "pand %[ftmp5], %[ftmp10], %[ftmp9] \n\t"
1493	1493	"psubb %[ftmp8], %[ftmp5], %[ftmp7] \n\t"
1494		− "and %[ftmp7], %[ftmp7], %[ftmp5] \n\t"
	1494	+ "pand %[ftmp7], %[ftmp7], %[ftmp5] \n\t"
1495	1495	"pavgb %[ftmp5], %[ftmp2], %[ftmp3] \n\t"
1496	1496	MMI_LDC1(%[ftmp11], %[addr1], 0x00)
1497	1497	"pavgb %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
1498		− "xor %[ftmp5], %[ftmp5], %[ftmp11] \n\t"
1499		− "and %[ftmp5], %[ftmp5], %[ff_pb_1] \n\t"
	1498	+ "pxor %[ftmp5], %[ftmp5], %[ftmp11] \n\t"
	1499	+ "pand %[ftmp5], %[ftmp5], %[ff_pb_1] \n\t"
1500	1500	"psubusb %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
1501	1501	"psubusb %[ftmp5], %[ftmp1], %[ftmp7] \n\t"
1502	1502	"paddusb %[ftmp7], %[ftmp7], %[ftmp1] \n\t"
@@ -1509,26 +1509,26 @@ void ff_deblock_v8_luma_8_mmi(uint8_t *pix, ptrdiff_t stride, int alpha, int bet · 6 unmodified lines
1509	1509	"psubusb %[ftmp4], %[ftmp4], %[ftmp6] \n\t"
1510	1510	"psubusb %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
1511	1511	"pcmpeqb %[ftmp7], %[ftmp7], %[ftmp4] \n\t"
1512		− "and %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
	1512	+ "pand %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
1513	1513	"psubb %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
1514		− "and %[ftmp6], %[ftmp9], %[ftmp7] \n\t"
	1514	+ "pand %[ftmp6], %[ftmp9], %[ftmp7] \n\t"
1515	1515	MMI_LDXC1(%[ftmp4], %[pix], %[stride], 0x00)
1516	1516	"pavgb %[ftmp7], %[ftmp2], %[ftmp3] \n\t"
1517	1517	MMI_LDXC1(%[ftmp11], %[pix], %[addr0], 0x00)
1518	1518	"pavgb %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1519		− "xor %[ftmp7], %[ftmp7], %[ftmp11] \n\t"
1520		− "and %[ftmp7], %[ftmp7], %[ff_pb_1] \n\t"
	1519	+ "pxor %[ftmp7], %[ftmp7], %[ftmp11] \n\t"
	1520	+ "pand %[ftmp7], %[ftmp7], %[ff_pb_1] \n\t"
1521	1521	"psubusb %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1522	1522	"psubusb %[ftmp7], %[ftmp4], %[ftmp6] \n\t"
1523	1523	"paddusb %[ftmp6], %[ftmp6], %[ftmp4] \n\t"
1524	1524	"pmaxub %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1525	1525	"pminub %[ftmp5], %[ftmp5], %[ftmp6] \n\t"
1526	1526	MMI_SDXC1(%[ftmp5], %[pix], %[stride], 0x00)
1527		− "xor %[ftmp6], %[ftmp2], %[ftmp3] \n\t"
	1527	+ "pxor %[ftmp6], %[ftmp2], %[ftmp3] \n\t"
1528	1528	"pcmpeqb %[ftmp5], %[ftmp5], %[ftmp5] \n\t"
1529		− "and %[ftmp6], %[ftmp6], %[ff_pb_1] \n\t"
1530		− "xor %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
1531		− "xor %[ftmp5], %[ftmp5], %[ftmp2] \n\t"
	1529	+ "pand %[ftmp6], %[ftmp6], %[ff_pb_1] \n\t"
	1530	+ "pxor %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
	1531	+ "pxor %[ftmp5], %[ftmp5], %[ftmp2] \n\t"
1532	1532	"pavgb %[ftmp4], %[ftmp4], %[ftmp1] \n\t"
1533	1533	"pavgb %[ftmp4], %[ftmp4], %[ff_pb_3] \n\t"
1534	1534	"pavgb %[ftmp5], %[ftmp5], %[ftmp3] \n\t"
@@ -1574,12 +1574,12 @@ static void deblock_v8_luma_intra_8_mmi(uint8_t *pix, ptrdiff_t stride, int alph · 39 unmodified lines
1574	1574
1575	1575	__asm__ volatile (
1576	1576	"ori %[tmp0], $0, 0x01 \n\t"
1577		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1577	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1578	1578	"mtc1 %[tmp0], %[ftmp9] \n\t"
1579	1579	PTR_SLL "%[addr0], %[stride], 0x02 \n\t"
1580	1580	PTR_ADDU "%[addr2], %[stride], %[stride] \n\t"
1581	1581	PTR_ADDIU "%[alpha], %[alpha], -0x01 \n\t"
1582		− PTR_SLL "%[ftmp11], %[ftmp9], %[ftmp9] \n\t"
	1582	+ "sslld %[ftmp11], %[ftmp9], %[ftmp9] \n\t"
1583	1583	"bltz %[alpha], 1f \n\t"
1584	1584	PTR_ADDU "%[addr1], %[addr2], %[stride] \n\t"
1585	1585	PTR_ADDIU "%[beta], %[beta], -0x01 \n\t"
@@ -1598,20 +1598,20 @@ static void deblock_v8_luma_intra_8_mmi(uint8_t *pix, ptrdiff_t stride, int alph · 12 unmodified lines
1598	1598	"psubusb %[ftmp7], %[ftmp3], %[ftmp2] \n\t"
1599	1599	"psubusb %[ftmp8], %[ftmp2], %[ftmp3] \n\t"
1600	1600	"packushb %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
1601		− "or %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
	1601	+ "por %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
1602	1602	MMI_SDC1(%[ftmp5], %[stack], 0x10)
1603	1603	"psubusb %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
1604	1604	"psubusb %[ftmp7], %[ftmp2], %[ftmp1] \n\t"
1605	1605	"psubusb %[ftmp5], %[ftmp1], %[ftmp2] \n\t"
1606		− "or %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
	1606	+ "por %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1607	1607	"psubusb %[ftmp5], %[ftmp5], %[ftmp6] \n\t"
1608		− "or %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
	1608	+ "por %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
1609	1609	"psubusb %[ftmp7], %[ftmp3], %[ftmp4] \n\t"
1610	1610	"psubusb %[ftmp5], %[ftmp4], %[ftmp3] \n\t"
1611		− "or %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
	1611	+ "por %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1612	1612	"psubusb %[ftmp5], %[ftmp5], %[ftmp6] \n\t"
1613		− "or %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
1614		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
	1613	+ "por %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
	1614	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
1615	1615	MMI_LDC1(%[ftmp5], %[stack], 0x10)
1616	1616	"pcmpeqb %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
1617	1617	"ldc1 %[ftmp10], %[ff_pb_1] \n\t"
@@ -1624,14 +1624,14 @@ static void deblock_v8_luma_intra_8_mmi(uint8_t *pix, ptrdiff_t stride, int alph · 6 unmodified lines
1624	1624	"psubusb %[ftmp7], %[ftmp7], %[ftmp5] \n\t"
1625	1625	MMI_LDC1(%[ftmp15], %[stack], 0x20)
1626	1626	"pcmpeqb %[ftmp7], %[ftmp7], %[ftmp8] \n\t"
1627		− "and %[ftmp7], %[ftmp7], %[ftmp15] \n\t"
	1627	+ "pand %[ftmp7], %[ftmp7], %[ftmp15] \n\t"
1628	1628	MMI_LDXC1(%[ftmp15], %[addr0], %[stride], 0x00)
1629	1629	"psubusb %[ftmp8], %[ftmp15], %[ftmp2] \n\t"
1630	1630	"psubusb %[ftmp5], %[ftmp2], %[ftmp15] \n\t"
1631	1631	"psubusb %[ftmp8], %[ftmp8], %[ftmp6] \n\t"
1632	1632	"psubusb %[ftmp5], %[ftmp5], %[ftmp6] \n\t"
1633	1633	"pcmpeqb %[ftmp5], %[ftmp5], %[ftmp8] \n\t"
1634		− "and %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
	1634	+ "pand %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1635	1635	MMI_LDXC1(%[ftmp14], %[pix], %[addr2], 0x00)
1636	1636	MMI_SDC1(%[ftmp5], %[stack], 0x30)
1637	1637	"psubusb %[ftmp8], %[ftmp14], %[ftmp3] \n\t"
@@ -1639,7 +1639,7 @@ static void deblock_v8_luma_intra_8_mmi(uint8_t *pix, ptrdiff_t stride, int alph · 1 unmodified line
1639	1639	"psubusb %[ftmp8], %[ftmp8], %[ftmp6] \n\t"
1640	1640	"psubusb %[ftmp5], %[ftmp5], %[ftmp6] \n\t"
1641	1641	"pcmpeqb %[ftmp5], %[ftmp5], %[ftmp8] \n\t"
1642		− "and %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
	1642	+ "pand %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1643	1643	MMI_SDC1(%[ftmp5], %[stack], 0x40)
1644	1644	"pavgb %[ftmp5], %[ftmp15], %[ftmp1] \n\t"
1645	1645	"pavgb %[ftmp6], %[ftmp2], %[ftmp3] \n\t"
@@ -1652,36 +1652,36 @@ static void deblock_v8_luma_intra_8_mmi(uint8_t *pix, ptrdiff_t stride, int alph · 6 unmodified lines
1652	1652	MMI_SDC1(%[ftmp7], %[stack], 0x00)
1653	1653	"psrlh %[ftmp7], %[ftmp7], %[ftmp9] \n\t"
1654	1654	"pavgb %[ftmp7], %[ftmp7], %[ftmp0] \n\t"
1655		− "xor %[ftmp7], %[ftmp7], %[ftmp5] \n\t"
1656		− "and %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
	1655	+ "pxor %[ftmp7], %[ftmp7], %[ftmp5] \n\t"
	1656	+ "pand %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
1657	1657	"psubb %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1658	1658	"pavgb %[ftmp6], %[ftmp15], %[ftmp4] \n\t"
1659	1659	"psubb %[ftmp7], %[ftmp15], %[ftmp4] \n\t"
1660	1660	"paddb %[ftmp8], %[ftmp8], %[ftmp8] \n\t"
1661	1661	"psubb %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
1662		− "and %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
	1662	+ "pand %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
1663	1663	"psubb %[ftmp6], %[ftmp6], %[ftmp7] \n\t"
1664	1664	MMI_LDC1(%[ftmp13], %[stack], 0x10)
1665	1665	"pavgb %[ftmp6], %[ftmp6], %[ftmp1] \n\t"
1666	1666	"psrlh %[ftmp8], %[ftmp8], %[ftmp11] \n\t"
1667	1667	"pavgb %[ftmp6], %[ftmp6], %[ftmp13] \n\t"
1668	1668	"pavgb %[ftmp8], %[ftmp8], %[ftmp0] \n\t"
1669		− "xor %[ftmp8], %[ftmp8], %[ftmp6] \n\t"
1670		− "and %[ftmp8], %[ftmp8], %[ftmp10] \n\t"
	1669	+ "pxor %[ftmp8], %[ftmp8], %[ftmp6] \n\t"
	1670	+ "pand %[ftmp8], %[ftmp8], %[ftmp10] \n\t"
1671	1671	"psubb %[ftmp6], %[ftmp6], %[ftmp8] \n\t"
1672		− "xor %[ftmp8], %[ftmp2], %[ftmp4] \n\t"
	1672	+ "pxor %[ftmp8], %[ftmp2], %[ftmp4] \n\t"
1673	1673	"pavgb %[ftmp7], %[ftmp2], %[ftmp4] \n\t"
1674		− "and %[ftmp8], %[ftmp8], %[ftmp10] \n\t"
	1674	+ "pand %[ftmp8], %[ftmp8], %[ftmp10] \n\t"
1675	1675	"psubb %[ftmp7], %[ftmp7], %[ftmp8] \n\t"
1676	1676	MMI_LDC1(%[ftmp13], %[stack], 0x30)
1677	1677	"pavgb %[ftmp7], %[ftmp7], %[ftmp1] \n\t"
1678	1678	MMI_LDC1(%[ftmp12], %[stack], 0x20)
1679		− "xor %[ftmp6], %[ftmp6], %[ftmp7] \n\t"
1680		− "xor %[ftmp7], %[ftmp7], %[ftmp2] \n\t"
1681		− "and %[ftmp6], %[ftmp6], %[ftmp13] \n\t"
1682		− "and %[ftmp7], %[ftmp7], %[ftmp12] \n\t"
1683		− "xor %[ftmp6], %[ftmp6], %[ftmp7] \n\t"
1684		− "xor %[ftmp6], %[ftmp6], %[ftmp2] \n\t"
	1679	+ "pxor %[ftmp6], %[ftmp6], %[ftmp7] \n\t"
	1680	+ "pxor %[ftmp7], %[ftmp7], %[ftmp2] \n\t"
	1681	+ "pand %[ftmp6], %[ftmp6], %[ftmp13] \n\t"
	1682	+ "pand %[ftmp7], %[ftmp7], %[ftmp12] \n\t"
	1683	+ "pxor %[ftmp6], %[ftmp6], %[ftmp7] \n\t"
	1684	+ "pxor %[ftmp6], %[ftmp6], %[ftmp2] \n\t"
1685	1685	MMI_SDXC1(%[ftmp6], %[addr0], %[addr1], 0x00)
1686	1686	MMI_LDC1(%[ftmp6], %[addr0], 0x00)
1687	1687	"paddb %[ftmp7], %[ftmp15], %[ftmp6] \n\t"
@@ -1692,16 +1692,16 @@ static void deblock_v8_luma_intra_8_mmi(uint8_t *pix, ptrdiff_t stride, int alph · 4 unmodified lines
1692	1692	"paddb %[ftmp7], %[ftmp7], %[ftmp12] \n\t"
1693	1693	"psrlh %[ftmp7], %[ftmp7], %[ftmp11] \n\t"
1694	1694	"pavgb %[ftmp7], %[ftmp7], %[ftmp0] \n\t"
1695		− "xor %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
1696		− "and %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
	1695	+ "pxor %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
	1696	+ "pand %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
1697	1697	MMI_LDC1(%[ftmp12], %[stack], 0x30)
1698	1698	"psubb %[ftmp6], %[ftmp6], %[ftmp7] \n\t"
1699		− "xor %[ftmp5], %[ftmp5], %[ftmp1] \n\t"
1700		− "xor %[ftmp6], %[ftmp6], %[ftmp15] \n\t"
1701		− "and %[ftmp5], %[ftmp5], %[ftmp12] \n\t"
1702		− "and %[ftmp6], %[ftmp6], %[ftmp12] \n\t"
1703		− "xor %[ftmp5], %[ftmp5], %[ftmp1] \n\t"
1704		− "xor %[ftmp6], %[ftmp6], %[ftmp15] \n\t"
	1699	+ "pxor %[ftmp5], %[ftmp5], %[ftmp1] \n\t"
	1700	+ "pxor %[ftmp6], %[ftmp6], %[ftmp15] \n\t"
	1701	+ "pand %[ftmp5], %[ftmp5], %[ftmp12] \n\t"
	1702	+ "pand %[ftmp6], %[ftmp6], %[ftmp12] \n\t"
	1703	+ "pxor %[ftmp5], %[ftmp5], %[ftmp1] \n\t"
	1704	+ "pxor %[ftmp6], %[ftmp6], %[ftmp15] \n\t"
1705	1705	MMI_SDXC1(%[ftmp5], %[addr0], %[addr2], 0x00)
1706	1706	MMI_SDXC1(%[ftmp6], %[addr0], %[stride], 0x00)
1707	1707	"pavgb %[ftmp5], %[ftmp14], %[ftmp4] \n\t"
@@ -1715,36 +1715,36 @@ static void deblock_v8_luma_intra_8_mmi(uint8_t *pix, ptrdiff_t stride, int alph · 7 unmodified lines
1715	1715	MMI_SDC1(%[ftmp7], %[stack], 0x00)
1716	1716	"psrlh %[ftmp7], %[ftmp7], %[ftmp9] \n\t"
1717	1717	"pavgb %[ftmp7], %[ftmp7], %[ftmp0] \n\t"
1718		− "xor %[ftmp7], %[ftmp7], %[ftmp5] \n\t"
1719		− "and %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
	1718	+ "pxor %[ftmp7], %[ftmp7], %[ftmp5] \n\t"
	1719	+ "pand %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
1720	1720	"psubb %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1721	1721	"pavgb %[ftmp6], %[ftmp14], %[ftmp1] \n\t"
1722	1722	"paddb %[ftmp8], %[ftmp8], %[ftmp8] \n\t"
1723	1723	"psubb %[ftmp7], %[ftmp14], %[ftmp1] \n\t"
1724	1724	"psubb %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
1725		− "and %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
	1725	+ "pand %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
1726	1726	"psubb %[ftmp6], %[ftmp6], %[ftmp7] \n\t"
1727	1727	MMI_LDC1(%[ftmp12], %[stack], 0x10)
1728	1728	"pavgb %[ftmp6], %[ftmp6], %[ftmp4] \n\t"
1729	1729	"pavgb %[ftmp6], %[ftmp6], %[ftmp12] \n\t"
1730	1730	"psrlh %[ftmp8], %[ftmp8], %[ftmp11] \n\t"
1731	1731	"pavgb %[ftmp8], %[ftmp8], %[ftmp0] \n\t"
1732		− "xor %[ftmp8], %[ftmp8], %[ftmp6] \n\t"
1733		− "and %[ftmp8], %[ftmp8], %[ftmp10] \n\t"
	1732	+ "pxor %[ftmp8], %[ftmp8], %[ftmp6] \n\t"
	1733	+ "pand %[ftmp8], %[ftmp8], %[ftmp10] \n\t"
1734	1734	"psubb %[ftmp6], %[ftmp6], %[ftmp8] \n\t"
1735		− "xor %[ftmp8], %[ftmp3], %[ftmp1] \n\t"
	1735	+ "pxor %[ftmp8], %[ftmp3], %[ftmp1] \n\t"
1736	1736	"pavgb %[ftmp7], %[ftmp3], %[ftmp1] \n\t"
1737		− "and %[ftmp8], %[ftmp8], %[ftmp10] \n\t"
	1737	+ "pand %[ftmp8], %[ftmp8], %[ftmp10] \n\t"
1738	1738	MMI_LDC1(%[ftmp12], %[stack], 0x40)
1739	1739	"psubb %[ftmp7], %[ftmp7], %[ftmp8] \n\t"
1740	1740	MMI_LDC1(%[ftmp13], %[stack], 0x20)
1741	1741	"pavgb %[ftmp7], %[ftmp7], %[ftmp4] \n\t"
1742		− "xor %[ftmp6], %[ftmp6], %[ftmp7] \n\t"
1743		− "xor %[ftmp7], %[ftmp7], %[ftmp3] \n\t"
1744		− "and %[ftmp6], %[ftmp6], %[ftmp12] \n\t"
1745		− "and %[ftmp7], %[ftmp7], %[ftmp13] \n\t"
1746		− "xor %[ftmp6], %[ftmp6], %[ftmp7] \n\t"
1747		− "xor %[ftmp6], %[ftmp6], %[ftmp3] \n\t"
	1742	+ "pxor %[ftmp6], %[ftmp6], %[ftmp7] \n\t"
	1743	+ "pxor %[ftmp7], %[ftmp7], %[ftmp3] \n\t"
	1744	+ "pand %[ftmp6], %[ftmp6], %[ftmp12] \n\t"
	1745	+ "pand %[ftmp7], %[ftmp7], %[ftmp13] \n\t"
	1746	+ "pxor %[ftmp6], %[ftmp6], %[ftmp7] \n\t"
	1747	+ "pxor %[ftmp6], %[ftmp6], %[ftmp3] \n\t"
1748	1748	MMI_SDC1(%[ftmp6], %[pix], 0x00)
1749	1749	MMI_LDXC1(%[ftmp6], %[pix], %[addr1], 0x00)
1750	1750	"paddb %[ftmp7], %[ftmp14], %[ftmp6] \n\t"
@@ -1755,16 +1755,16 @@ static void deblock_v8_luma_intra_8_mmi(uint8_t *pix, ptrdiff_t stride, int alph · 4 unmodified lines
1755	1755	"paddb %[ftmp7], %[ftmp7], %[ftmp12] \n\t"
1756	1756	"psrlh %[ftmp7], %[ftmp7], %[ftmp11] \n\t"
1757	1757	"pavgb %[ftmp7], %[ftmp7], %[ftmp0] \n\t"
1758		− "xor %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
1759		− "and %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
	1758	+ "pxor %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
	1759	+ "pand %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
1760	1760	MMI_LDC1(%[ftmp12], %[stack], 0x40)
1761	1761	"psubb %[ftmp6], %[ftmp6], %[ftmp7] \n\t"
1762		− "xor %[ftmp5], %[ftmp5], %[ftmp4] \n\t"
1763		− "xor %[ftmp6], %[ftmp6], %[ftmp14] \n\t"
1764		− "and %[ftmp5], %[ftmp5], %[ftmp12] \n\t"
1765		− "and %[ftmp6], %[ftmp6], %[ftmp12] \n\t"
1766		− "xor %[ftmp5], %[ftmp5], %[ftmp4] \n\t"
1767		− "xor %[ftmp6], %[ftmp6], %[ftmp14] \n\t"
	1762	+ "pxor %[ftmp5], %[ftmp5], %[ftmp4] \n\t"
	1763	+ "pxor %[ftmp6], %[ftmp6], %[ftmp14] \n\t"
	1764	+ "pand %[ftmp5], %[ftmp5], %[ftmp12] \n\t"
	1765	+ "pand %[ftmp6], %[ftmp6], %[ftmp12] \n\t"
	1766	+ "pxor %[ftmp5], %[ftmp5], %[ftmp4] \n\t"
	1767	+ "pxor %[ftmp6], %[ftmp6], %[ftmp14] \n\t"
1768	1768	MMI_SDXC1(%[ftmp5], %[pix], %[stride], 0x00)
1769	1769	MMI_SDXC1(%[ftmp6], %[pix], %[addr2], 0x00)
1770	1770	"1: \n\t"
@@ -1808,7 +1808,7 @@ void ff_deblock_v_chroma_8_mmi(uint8_t *pix, ptrdiff_t stride, int alpha, · 37 unmodified lines
1808	1808	MMI_LDC1(%[ftmp3], %[pix], 0x00)
1809	1809	MMI_LDXC1(%[ftmp4], %[pix], %[stride], 0x00)
1810	1810
1811		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1811	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1812	1812	"mtc1 %[alpha], %[ftmp5] \n\t"
1813	1813	"mtc1 %[beta], %[ftmp6] \n\t"
1814	1814	"pshufh %[ftmp5], %[ftmp5], %[ftmp0] \n\t"
@@ -1817,29 +1817,29 @@ void ff_deblock_v_chroma_8_mmi(uint8_t *pix, ptrdiff_t stride, int alpha, · 2 unmodified lines
1817	1817	"packushb %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
1818	1818	"psubusb %[ftmp7], %[ftmp3], %[ftmp2] \n\t"
1819	1819	"psubusb %[ftmp8], %[ftmp2], %[ftmp3] \n\t"
1820		− "or %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
	1820	+ "por %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
1821	1821	"psubusb %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
1822	1822	"psubusb %[ftmp7], %[ftmp2], %[ftmp1] \n\t"
1823	1823	"psubusb %[ftmp5], %[ftmp1], %[ftmp2] \n\t"
1824		− "or %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
	1824	+ "por %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1825	1825	"psubusb %[ftmp5], %[ftmp5], %[ftmp6] \n\t"
1826		− "or %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
	1826	+ "por %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
1827	1827	"psubusb %[ftmp7], %[ftmp3], %[ftmp4] \n\t"
1828	1828	"psubusb %[ftmp5], %[ftmp4], %[ftmp3] \n\t"
1829		− "or %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
	1829	+ "por %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1830	1830	"psubusb %[ftmp5], %[ftmp5], %[ftmp6] \n\t"
1831		− "or %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
1832		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
	1831	+ "por %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
	1832	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
1833	1833	"pcmpeqb %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
1834	1834	MMI_ULWC1(%[ftmp7], %[tc0], 0x00)
1835	1835	"punpcklbh %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
1836		− "and %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
	1836	+ "pand %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
1837	1837	"pcmpeqb %[ftmp5], %[ftmp5], %[ftmp5] \n\t"
1838		− "xor %[ftmp6], %[ftmp2], %[ftmp3] \n\t"
1839		− "xor %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
1840		− "and %[ftmp6], %[ftmp6], %[ff_pb_1] \n\t"
	1838	+ "pxor %[ftmp6], %[ftmp2], %[ftmp3] \n\t"
	1839	+ "pxor %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
	1840	+ "pand %[ftmp6], %[ftmp6], %[ff_pb_1] \n\t"
1841	1841	"pavgb %[ftmp4], %[ftmp4], %[ftmp1] \n\t"
1842		− "xor %[ftmp5], %[ftmp5], %[ftmp2] \n\t"
	1842	+ "pxor %[ftmp5], %[ftmp5], %[ftmp2] \n\t"
1843	1843	"pavgb %[ftmp4], %[ftmp4], %[ff_pb_3] \n\t"
1844	1844	"pavgb %[ftmp5], %[ftmp5], %[ftmp3] \n\t"
1845	1845	"pavgb %[ftmp4], %[ftmp4], %[ftmp6] \n\t"
@@ -1891,7 +1891,7 @@ void ff_deblock_v_chroma_intra_8_mmi(uint8_t *pix, ptrdiff_t stride, int alpha, · 45 unmodified lines
1891	1891	MMI_LDC1(%[ftmp3], %[pix], 0x00)
1892	1892	MMI_LDXC1(%[ftmp4], %[pix], %[stride], 0x00)
1893	1893
1894		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1894	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1895	1895	"mtc1 %[alpha], %[ftmp5] \n\t"
1896	1896	"mtc1 %[beta], %[ftmp6] \n\t"
1897	1897	"pshufh %[ftmp5], %[ftmp5], %[ftmp0] \n\t"
@@ -1900,36 +1900,36 @@ void ff_deblock_v_chroma_intra_8_mmi(uint8_t *pix, ptrdiff_t stride, int alpha, · 2 unmodified lines
1900	1900	"packushb %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
1901	1901	"psubusb %[ftmp7], %[ftmp3], %[ftmp2] \n\t"
1902	1902	"psubusb %[ftmp8], %[ftmp2], %[ftmp3] \n\t"
1903		− "or %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
	1903	+ "por %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
1904	1904	"psubusb %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
1905	1905	"psubusb %[ftmp7], %[ftmp2], %[ftmp1] \n\t"
1906	1906	"psubusb %[ftmp5], %[ftmp1], %[ftmp2] \n\t"
1907		− "or %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
	1907	+ "por %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1908	1908	"psubusb %[ftmp5], %[ftmp5], %[ftmp6] \n\t"
1909		− "or %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
	1909	+ "por %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
1910	1910	"psubusb %[ftmp7], %[ftmp3], %[ftmp4] \n\t"
1911	1911	"psubusb %[ftmp5], %[ftmp4], %[ftmp3] \n\t"
1912		− "or %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
	1912	+ "por %[ftmp5], %[ftmp5], %[ftmp7] \n\t"
1913	1913	"psubusb %[ftmp5], %[ftmp5], %[ftmp6] \n\t"
1914		− "or %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
1915		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
	1914	+ "por %[ftmp8], %[ftmp8], %[ftmp5] \n\t"
	1915	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
1916	1916	"pcmpeqb %[ftmp8], %[ftmp8], %[ftmp7] \n\t"
1917	1917	"mov.d %[ftmp6], %[ftmp2] \n\t"
1918	1918	"mov.d %[ftmp7], %[ftmp3] \n\t"
1919		− "xor %[ftmp5], %[ftmp2], %[ftmp4] \n\t"
1920		− "and %[ftmp5], %[ftmp5], %[ff_pb_1] \n\t"
	1919	+ "pxor %[ftmp5], %[ftmp2], %[ftmp4] \n\t"
	1920	+ "pand %[ftmp5], %[ftmp5], %[ff_pb_1] \n\t"
1921	1921	"pavgb %[ftmp2], %[ftmp2], %[ftmp4] \n\t"
1922	1922	"psubusb %[ftmp2], %[ftmp2], %[ftmp5] \n\t"
1923	1923	"pavgb %[ftmp2], %[ftmp2], %[ftmp1] \n\t"
1924		− "xor %[ftmp5], %[ftmp3], %[ftmp1] \n\t"
1925		− "and %[ftmp5], %[ftmp5], %[ff_pb_1] \n\t"
	1924	+ "pxor %[ftmp5], %[ftmp3], %[ftmp1] \n\t"
	1925	+ "pand %[ftmp5], %[ftmp5], %[ff_pb_1] \n\t"
1926	1926	"pavgb %[ftmp3], %[ftmp3], %[ftmp1] \n\t"
1927	1927	"psubusb %[ftmp3], %[ftmp3], %[ftmp5] \n\t"
1928	1928	"pavgb %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
1929	1929	"psubb %[ftmp2], %[ftmp2], %[ftmp6] \n\t"
1930	1930	"psubb %[ftmp3], %[ftmp3], %[ftmp7] \n\t"
1931		− "and %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
1932		− "and %[ftmp3], %[ftmp3], %[ftmp8] \n\t"
	1931	+ "pand %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
	1932	+ "pand %[ftmp3], %[ftmp3], %[ftmp8] \n\t"
1933	1933	"paddb %[ftmp2], %[ftmp2], %[ftmp6] \n\t"
1934	1934	"paddb %[ftmp3], %[ftmp3], %[ftmp7] \n\t"
1935	1935
@@ -1996,7 +1996,7 @@ void ff_deblock_h_chroma_8_mmi(uint8_t *pix, ptrdiff_t stride, int alpha, int be · 60 unmodified lines
1996	1996	"mov.d %[ftmp9], %[ftmp0] \n\t"
1997	1997	"mov.d %[ftmp10], %[ftmp3] \n\t"
1998	1998
1999		− "xor %[ftmp8], %[ftmp8], %[ftmp8] \n\t"
	1999	+ "pxor %[ftmp8], %[ftmp8], %[ftmp8] \n\t"
2000	2000	"mtc1 %[alpha], %[ftmp4] \n\t"
2001	2001	"mtc1 %[beta], %[ftmp5] \n\t"
2002	2002	"pshufh %[ftmp4], %[ftmp4], %[ftmp8] \n\t"
@@ -2005,29 +2005,29 @@ void ff_deblock_h_chroma_8_mmi(uint8_t *pix, ptrdiff_t stride, int alpha, int be · 2 unmodified lines
2005	2005	"packushb %[ftmp5], %[ftmp5], %[ftmp5] \n\t"
2006	2006	"psubusb %[ftmp6], %[ftmp2], %[ftmp1] \n\t"
2007	2007	"psubusb %[ftmp7], %[ftmp1], %[ftmp2] \n\t"
2008		− "or %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
	2008	+ "por %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
2009	2009	"psubusb %[ftmp7], %[ftmp7], %[ftmp4] \n\t"
2010	2010	"psubusb %[ftmp6], %[ftmp1], %[ftmp0] \n\t"
2011	2011	"psubusb %[ftmp4], %[ftmp0], %[ftmp1] \n\t"
2012		− "or %[ftmp4], %[ftmp4], %[ftmp6] \n\t"
	2012	+ "por %[ftmp4], %[ftmp4], %[ftmp6] \n\t"
2013	2013	"psubusb %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
2014		− "or %[ftmp7], %[ftmp7], %[ftmp4] \n\t"
	2014	+ "por %[ftmp7], %[ftmp7], %[ftmp4] \n\t"
2015	2015	"psubusb %[ftmp6], %[ftmp2], %[ftmp3] \n\t"
2016	2016	"psubusb %[ftmp4], %[ftmp3], %[ftmp2] \n\t"
2017		− "or %[ftmp4], %[ftmp4], %[ftmp6] \n\t"
	2017	+ "por %[ftmp4], %[ftmp4], %[ftmp6] \n\t"
2018	2018	"psubusb %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
2019		− "or %[ftmp7], %[ftmp7], %[ftmp4] \n\t"
2020		− "xor %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
	2019	+ "por %[ftmp7], %[ftmp7], %[ftmp4] \n\t"
	2020	+ "pxor %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
2021	2021	"pcmpeqb %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
2022	2022	MMI_ULWC1(%[ftmp6], %[tc0], 0x00)
2023	2023	"punpcklbh %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
2024		− "and %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
	2024	+ "pand %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
2025	2025	"pcmpeqb %[ftmp4], %[ftmp4], %[ftmp4] \n\t"
2026		− "xor %[ftmp5], %[ftmp1], %[ftmp2] \n\t"
2027		− "xor %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
2028		− "and %[ftmp5], %[ftmp5], %[ff_pb_1] \n\t"
	2026	+ "pxor %[ftmp5], %[ftmp1], %[ftmp2] \n\t"
	2027	+ "pxor %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
	2028	+ "pand %[ftmp5], %[ftmp5], %[ff_pb_1] \n\t"
2029	2029	"pavgb %[ftmp3], %[ftmp3], %[ftmp0] \n\t"
2030		− "xor %[ftmp4], %[ftmp4], %[ftmp1] \n\t"
	2030	+ "pxor %[ftmp4], %[ftmp4], %[ftmp1] \n\t"
2031	2031	"pavgb %[ftmp3], %[ftmp3], %[ff_pb_3] \n\t"
2032	2032	"pavgb %[ftmp4], %[ftmp4], %[ftmp2] \n\t"
2033	2033	"pavgb %[ftmp3], %[ftmp3], %[ftmp5] \n\t"
@@ -2134,7 +2134,7 @@ void ff_deblock_h_chroma_intra_8_mmi(uint8_t *pix, ptrdiff_t stride, int alpha, · 100 unmodified lines
2134	2134	"punpcklwd %[ftmp0], %[ftmp0], %[ftmp4] \n\t"
2135	2135	"punpcklwd %[ftmp2], %[ftmp2], %[ftmp6] \n\t"
2136	2136
2137		− "xor %[ftmp8], %[ftmp8], %[ftmp8] \n\t"
	2137	+ "pxor %[ftmp8], %[ftmp8], %[ftmp8] \n\t"
2138	2138	"mtc1 %[alpha], %[ftmp4] \n\t"
2139	2139	"mtc1 %[beta], %[ftmp5] \n\t"
2140	2140	"pshufh %[ftmp4], %[ftmp4], %[ftmp8] \n\t"
@@ -2143,36 +2143,36 @@ void ff_deblock_h_chroma_intra_8_mmi(uint8_t *pix, ptrdiff_t stride, int alpha, · 2 unmodified lines
2143	2143	"packushb %[ftmp5], %[ftmp5], %[ftmp5] \n\t"
2144	2144	"psubusb %[ftmp6], %[ftmp2], %[ftmp1] \n\t"
2145	2145	"psubusb %[ftmp7], %[ftmp1], %[ftmp2] \n\t"
2146		− "or %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
	2146	+ "por %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
2147	2147	"psubusb %[ftmp7], %[ftmp7], %[ftmp4] \n\t"
2148	2148	"psubusb %[ftmp6], %[ftmp1], %[ftmp0] \n\t"
2149	2149	"psubusb %[ftmp4], %[ftmp0], %[ftmp1] \n\t"
2150		− "or %[ftmp4], %[ftmp4], %[ftmp6] \n\t"
	2150	+ "por %[ftmp4], %[ftmp4], %[ftmp6] \n\t"
2151	2151	"psubusb %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
2152		− "or %[ftmp7], %[ftmp7], %[ftmp4] \n\t"
	2152	+ "por %[ftmp7], %[ftmp7], %[ftmp4] \n\t"
2153	2153	"psubusb %[ftmp6], %[ftmp2], %[ftmp3] \n\t"
2154	2154	"psubusb %[ftmp4], %[ftmp3], %[ftmp2] \n\t"
2155		− "or %[ftmp4], %[ftmp4], %[ftmp6] \n\t"
	2155	+ "por %[ftmp4], %[ftmp4], %[ftmp6] \n\t"
2156	2156	"psubusb %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
2157		− "or %[ftmp7], %[ftmp7], %[ftmp4] \n\t"
2158		− "xor %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
	2157	+ "por %[ftmp7], %[ftmp7], %[ftmp4] \n\t"
	2158	+ "pxor %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
2159	2159	"pcmpeqb %[ftmp7], %[ftmp7], %[ftmp6] \n\t"
2160	2160	"mov.d %[ftmp5], %[ftmp1] \n\t"
2161	2161	"mov.d %[ftmp6], %[ftmp2] \n\t"
2162		− "xor %[ftmp4], %[ftmp1], %[ftmp3] \n\t"
2163		− "and %[ftmp4], %[ftmp4], %[ff_pb_1] \n\t"
	2162	+ "pxor %[ftmp4], %[ftmp1], %[ftmp3] \n\t"
	2163	+ "pand %[ftmp4], %[ftmp4], %[ff_pb_1] \n\t"
2164	2164	"pavgb %[ftmp1], %[ftmp1], %[ftmp3] \n\t"
2165	2165	"psubusb %[ftmp1], %[ftmp1], %[ftmp4] \n\t"
2166	2166	"pavgb %[ftmp1], %[ftmp1], %[ftmp0] \n\t"
2167		− "xor %[ftmp4], %[ftmp2], %[ftmp0] \n\t"
2168		− "and %[ftmp4], %[ftmp4], %[ff_pb_1] \n\t"
	2167	+ "pxor %[ftmp4], %[ftmp2], %[ftmp0] \n\t"
	2168	+ "pand %[ftmp4], %[ftmp4], %[ff_pb_1] \n\t"
2169	2169	"pavgb %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
2170	2170	"psubusb %[ftmp2], %[ftmp2], %[ftmp4] \n\t"
2171	2171	"pavgb %[ftmp2], %[ftmp2], %[ftmp3] \n\t"
2172	2172	"psubb %[ftmp1], %[ftmp1], %[ftmp5] \n\t"
2173	2173	"psubb %[ftmp2], %[ftmp2], %[ftmp6] \n\t"
2174		− "and %[ftmp1], %[ftmp1], %[ftmp7] \n\t"
2175		− "and %[ftmp2], %[ftmp2], %[ftmp7] \n\t"
	2174	+ "pand %[ftmp1], %[ftmp1], %[ftmp7] \n\t"
	2175	+ "pand %[ftmp2], %[ftmp2], %[ftmp7] \n\t"
2176	2176	"paddb %[ftmp1], %[ftmp1], %[ftmp5] \n\t"
2177	2177	"paddb %[ftmp2], %[ftmp2], %[ftmp6] \n\t"
2178	2178

libavcodec/mips/h264pred_mmi.c+9−9

@@ -162,7 +162,7 @@ void ff_pred8x8l_top_dc_8_mmi(uint8_t *src, int has_topleft, · 161 unmodified lines
162	162	DECLARE_VAR_ADDRT;
163	163
164	164	__asm__ volatile (
165		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	165	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
166	166	MMI_ULDC1(%[ftmp10], %[srcA], 0x00)
167	167	MMI_ULDC1(%[ftmp9], %[src0], 0x00)
168	168	MMI_ULDC1(%[ftmp8], %[src1], 0x00)
@@ -266,7 +266,7 @@ void ff_pred8x8l_dc_8_mmi(uint8_t *src, int has_topleft, int has_topright, · 97 unmodified lines
266	266	MMI_ULDC1(%[ftmp4], %[srcA], 0x00)
267	267	MMI_ULDC1(%[ftmp5], %[src0], 0x00)
268	268	MMI_ULDC1(%[ftmp6], %[src1], 0x00)
269		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	269	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
270	270	"dli %[tmp0], 0x03 \n\t"
271	271	"punpcklbh %[ftmp7], %[ftmp4], %[ftmp0] \n\t"
272	272	"punpckhbh %[ftmp8], %[ftmp4], %[ftmp0] \n\t"
@@ -357,7 +357,7 @@ void ff_pred8x8l_vertical_8_mmi(uint8_t *src, int has_topleft, · 84 unmodified lines
357	357	DECLARE_VAR_ALL64;
358	358
359	359	__asm__ volatile (
360		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	360	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
361	361	MMI_LDC1(%[ftmp3], %[srcA], 0x00)
362	362	MMI_LDC1(%[ftmp4], %[src0], 0x00)
363	363	MMI_LDC1(%[ftmp5], %[src1], 0x00)
@@ -530,7 +530,7 @@ void ff_pred8x8_top_dc_8_mmi(uint8_t *src, ptrdiff_t stride) · 166 unmodified lines
530	530
531	531	__asm__ volatile (
532	532	"dli %[tmp0], 0x02 \n\t"
533		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	533	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
534	534	PTR_SUBU "%[addr0], %[src], %[stride] \n\t"
535	535	MMI_LDC1(%[ftmp1], %[addr0], 0x00)
536	536	"punpcklbh %[ftmp2], %[ftmp1], %[ftmp0] \n\t"
@@ -640,7 +640,7 @@ void ff_pred8x8_dc_8_mmi(uint8_t *src, ptrdiff_t stride) · 103 unmodified lines
640	640	PTR_SRL "%[addr4], 0x02 \n\t"
641	641	PTR_SRL "%[addr1], 0x02 \n\t"
642	642	PTR_SRL "%[addr2], 0x03 \n\t"
643		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	643	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
644	644	"dmtc1 %[addr3], %[ftmp1] \n\t"
645	645	"pshufh %[ftmp1], %[ftmp1], %[ftmp0] \n\t"
646	646	"dmtc1 %[addr4], %[ftmp2] \n\t"
@@ -757,9 +757,9 @@ static inline void pred16x16_plane_compat_mmi(uint8_t *src, int stride, · 110 unmodified lines
757	757	"dmtc1 %[tmp0], %[ftmp4] \n\t"
758	758	MMI_ULDC1(%[ftmp0], %[addr0], -0x01)
759	759	MMI_ULDC1(%[ftmp2], %[addr0], 0x08)
760		− "dsrl %[ftmp1], %[ftmp0], %[ftmp4] \n\t"
761		− "dsrl %[ftmp3], %[ftmp2], %[ftmp4] \n\t"
762		− "xor %[ftmp4], %[ftmp4], %[ftmp4] \n\t"
	760	+ "ssrld %[ftmp1], %[ftmp0], %[ftmp4] \n\t"
	761	+ "ssrld %[ftmp3], %[ftmp2], %[ftmp4] \n\t"
	762	+ "pxor %[ftmp4], %[ftmp4], %[ftmp4] \n\t"
763	763	"punpcklbh %[ftmp0], %[ftmp0], %[ftmp4] \n\t"
764	764	"punpcklbh %[ftmp1], %[ftmp1], %[ftmp4] \n\t"
765	765	"punpcklbh %[ftmp2], %[ftmp2], %[ftmp4] \n\t"
@@ -915,7 +915,7 @@ static inline void pred16x16_plane_compat_mmi(uint8_t *src, int stride, · 149 unmodified lines
915	915	"dmul %[tmp3], %[tmp3], %[tmp2] \n\t"
916	916	"dsubu %[tmp5], %[tmp5], %[tmp3] \n\t"
917	917
918		− "xor %[ftmp4], %[ftmp4], %[ftmp4] \n\t"
	918	+ "pxor %[ftmp4], %[ftmp4], %[ftmp4] \n\t"
919	919	"dmtc1 %[tmp0], %[ftmp0] \n\t"
920	920	"pshufh %[ftmp0], %[ftmp0], %[ftmp4] \n\t"
921	921	"dmtc1 %[tmp1], %[ftmp5] \n\t"

libavcodec/mips/h264qpel_mmi.c+13−13

@@ -114,7 +114,7 @@ static void put_h264_qpel4_h_lowpass_mmi(uint8_t dst, const uint8_t src, · 113 unmodified lines
114	114	DECLARE_VAR_LOW32;
115	115
116	116	__asm__ volatile (
117		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	117	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
118	118	"dli %[tmp0], 0x04 \n\t"
119	119	"1: \n\t"
120	120	MMI_ULWC1(%[ftmp1], %[src], -0x02)
@@ -169,7 +169,7 @@ static void put_h264_qpel8_h_lowpass_mmi(uint8_t dst, const uint8_t src, · 48 unmodified lines
169	169	DECLARE_VAR_ALL64;
170	170
171	171	__asm__ volatile (
172		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	172	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
173	173	"dli %[tmp0], 0x08 \n\t"
174	174	"1: \n\t"
175	175	MMI_ULDC1(%[ftmp1], %[src], -0x02)
@@ -250,7 +250,7 @@ static void avg_h264_qpel4_h_lowpass_mmi(uint8_t dst, const uint8_t src, · 74 unmodified lines
250	250	DECLARE_VAR_LOW32;
251	251
252	252	__asm__ volatile (
253		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	253	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
254	254	"dli %[tmp0], 0x04 \n\t"
255	255	"1: \n\t"
256	256	MMI_ULWC1(%[ftmp1], %[src], -0x02)
@@ -307,7 +307,7 @@ static void avg_h264_qpel8_h_lowpass_mmi(uint8_t dst, const uint8_t src, · 50 unmodified lines
307	307	DECLARE_VAR_ALL64;
308	308
309	309	__asm__ volatile (
310		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	310	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
311	311	"dli %[tmp0], 0x08 \n\t"
312	312	"1: \n\t"
313	313	MMI_ULDC1(%[ftmp1], %[src], -0x02)
@@ -394,7 +394,7 @@ static void put_h264_qpel4_v_lowpass_mmi(uint8_t dst, const uint8_t src, · 80 unmodified lines
394	394	__asm__ volatile (
395	395	".set push \n\t"
396	396	".set noreorder \n\t"
397		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	397	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
398	398	"dli %[tmp0], 0x02 \n\t"
399	399	MMI_LWC1(%[ftmp1], %[src], 0x00)
400	400	"mtc1 %[tmp0], %[ftmp10] \n\t"
@@ -516,7 +516,7 @@ static void put_h264_qpel8_v_lowpass_mmi(uint8_t dst, const uint8_t src, · 115 unmodified lines
516	516	PTR_ADDU "%[src], %[src], %[srcStride] \n\t"
517	517	MMI_LWC1(%[ftmp2], %[src], 0x00)
518	518	PTR_ADDU "%[src], %[src], %[srcStride] \n\t"
519		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
	519	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
520	520	MMI_LWC1(%[ftmp3], %[src], 0x00)
521	521	PTR_ADDU "%[src], %[src], %[srcStride] \n\t"
522	522	MMI_LWC1(%[ftmp4], %[src], 0x00)
@@ -812,7 +812,7 @@ static void avg_h264_qpel4_v_lowpass_mmi(uint8_t dst, const uint8_t src, · 289 unmodified lines
812	812	".set push \n\t"
813	813	".set noreorder \n\t"
814	814	"dli %[tmp0], 0x02 \n\t"
815		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
	815	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
816	816	"mtc1 %[tmp0], %[ftmp9] \n\t"
817	817	"dli %[tmp0], 0x05 \n\t"
818	818	MMI_LWC1(%[ftmp0], %[src], 0x00)
@@ -930,7 +930,7 @@ static void avg_h264_qpel8_v_lowpass_mmi(uint8_t dst, const uint8_t src, · 111 unmodified lines
930	930	".set push \n\t"
931	931	".set noreorder \n\t"
932	932	"dli %[tmp0], 0x02 \n\t"
933		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
	933	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
934	934	"mtc1 %[tmp0], %[ftmp9] \n\t"
935	935	"dli %[tmp0], 0x05 \n\t"
936	936	MMI_LWC1(%[ftmp0], %[src], 0x00)
@@ -1269,7 +1269,7 @@ static void put_h264_qpel4_hv_lowpass_mmi(uint8_t dst, const uint8_t src, · 332 unmodified lines
1269	1269	src -= 2*srcStride;
1270	1270
1271	1271	__asm__ volatile (
1272		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1272	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1273	1273	"dli %[tmp0], 0x09 \n\t"
1274	1274	"1: \n\t"
1275	1275	MMI_ULWC1(%[ftmp1], %[src], -0x02)
@@ -1347,7 +1347,7 @@ static void put_h264_qpel8or16_hv1_lowpass_mmi(int16_t *tmp, · 71 unmodified lines
1347	1347	MMI_ULWC1(%[ftmp0], %[src], 0x00)
1348	1348	"mtc1 %[tmp0], %[ftmp10] \n\t"
1349	1349	PTR_ADDU "%[src], %[src], %[srcStride] \n\t"
1350		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
	1350	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
1351	1351	MMI_ULWC1(%[ftmp1], %[src], 0x00)
1352	1352	PTR_ADDU "%[src], %[src], %[srcStride] \n\t"
1353	1353	MMI_ULWC1(%[ftmp2], %[src], 0x00)
@@ -1684,7 +1684,7 @@ static void put_h264_qpel8_h_lowpass_l2_mmi(uint8_t dst, const uint8_t src, · 330 unmodified lines
1684	1684	"dli %[tmp0], 0x02 \n\t"
1685	1685	"mtc1 %[tmp0], %[ftmp7] \n\t"
1686	1686	"dli %[tmp0], 0x05 \n\t"
1687		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1687	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1688	1688	"mtc1 %[tmp0], %[ftmp8] \n\t"
1689	1689	"1: \n\t"
1690	1690	MMI_ULDC1(%[ftmp1], %[src], 0x00)
@@ -1833,7 +1833,7 @@ static void avg_h264_qpel4_hv_lowpass_mmi(uint8_t dst, const uint8_t src, · 142 unmodified lines
1833	1833	src -= 2*srcStride;
1834	1834
1835	1835	__asm__ volatile (
1836		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1836	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1837	1837	"dli %[tmp0], 0x09 \n\t"
1838	1838	"1: \n\t"
1839	1839	MMI_ULWC1(%[ftmp1], %[src], -0x02)
@@ -2005,7 +2005,7 @@ static void avg_h264_qpel8_h_lowpass_l2_mmi(uint8_t dst, const uint8_t src, · 165 unmodified lines
2005	2005	"ori %[tmp0], $0, 0x8 \n\t"
2006	2006	"mtc1 %[tmp1], %[ftmp7] \n\t"
2007	2007	"dli %[tmp1], 0x05 \n\t"
2008		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	2008	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
2009	2009	"mtc1 %[tmp1], %[ftmp8] \n\t"
2010	2010	"1: \n\t"
2011	2011	MMI_ULDC1(%[ftmp1], %[src], 0x00)

libavcodec/mips/hevcdsp_mmi.c+16−16

@@ -46,7 +46,7 @@ void ff_hevc_put_hevc_qpel_h##w##_8_mmi(int16_t dst, uint8_t _src, \ · 45 unmodified lines
46	46	"punpcklbh %[ftmp1], %[ftmp0], %[ftmp1] \n\t" \
47	47	"psrah %[ftmp1], %[ftmp1], %[ftmp0] \n\t" \
48	48	"psrah %[ftmp2], %[ftmp2], %[ftmp0] \n\t" \
49		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
	49	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
50	50	\
51	51	"1: \n\t" \
52	52	"2: \n\t" \
@@ -147,7 +147,7 @@ void ff_hevc_put_hevc_qpel_hv##w##_8_mmi(int16_t dst, uint8_t _src, \ · 94 unmodified lines
147	147	"punpcklbh %[ftmp1], %[ftmp0], %[ftmp1] \n\t" \
148	148	"psrah %[ftmp1], %[ftmp1], %[ftmp0] \n\t" \
149	149	"psrah %[ftmp2], %[ftmp2], %[ftmp0] \n\t" \
150		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
	150	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
151	151	\
152	152	"1: \n\t" \
153	153	"2: \n\t" \
@@ -344,7 +344,7 @@ void ff_hevc_put_hevc_qpel_bi_h##w##_8_mmi(uint8_t *_dst, \ · 190 unmodified lines
344	344	"punpcklbh %[ftmp1], %[ftmp0], %[ftmp1] \n\t" \
345	345	"psrah %[ftmp1], %[ftmp1], %[ftmp0] \n\t" \
346	346	"psrah %[ftmp2], %[ftmp2], %[ftmp0] \n\t" \
347		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
	347	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
348	348	"punpcklhw %[offset], %[offset], %[offset] \n\t" \
349	349	"punpcklwd %[offset], %[offset], %[offset] \n\t" \
350	350	\
@@ -403,7 +403,7 @@ void ff_hevc_put_hevc_qpel_bi_h##w##_8_mmi(uint8_t *_dst, \ · 52 unmodified lines
403	403	"psraw %[ftmp6], %[ftmp6], %[shift] \n\t" \
404	404	"packsswh %[ftmp5], %[ftmp5], %[ftmp6] \n\t" \
405	405	"pcmpgth %[ftmp7], %[ftmp5], %[ftmp0] \n\t" \
406		− "and %[ftmp3], %[ftmp5], %[ftmp7] \n\t" \
	406	+ "pand %[ftmp3], %[ftmp5], %[ftmp7] \n\t" \
407	407	"packushb %[ftmp3], %[ftmp3], %[ftmp3] \n\t" \
408	408	"gsswlc1 %[ftmp3], 0x03(%[dst]) \n\t" \
409	409	"gsswrc1 %[ftmp3], 0x00(%[dst]) \n\t" \
@@ -480,7 +480,7 @@ void ff_hevc_put_hevc_qpel_bi_hv##w##_8_mmi(uint8_t *_dst, \ · 70 unmodified lines
480	480	"punpcklbh %[ftmp1], %[ftmp0], %[ftmp1] \n\t" \
481	481	"psrah %[ftmp1], %[ftmp1], %[ftmp0] \n\t" \
482	482	"psrah %[ftmp2], %[ftmp2], %[ftmp0] \n\t" \
483		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
	483	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
484	484	\
485	485	"1: \n\t" \
486	486	"2: \n\t" \
@@ -612,7 +612,7 @@ void ff_hevc_put_hevc_qpel_bi_hv##w##_8_mmi(uint8_t *_dst, \ · 125 unmodified lines
612	612	"packsswh %[ftmp3], %[ftmp3], %[ftmp5] \n\t" \
613	613	"gsldlc1 %[ftmp4], 0x07(%[src2]) \n\t" \
614	614	"gsldrc1 %[ftmp4], 0x00(%[src2]) \n\t" \
615		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t" \
	615	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t" \
616	616	"li %[rtmp0], 0x10 \n\t" \
617	617	"dmtc1 %[rtmp0], %[ftmp8] \n\t" \
618	618	"punpcklhw %[ftmp5], %[ftmp7], %[ftmp3] \n\t" \
@@ -631,7 +631,7 @@ void ff_hevc_put_hevc_qpel_bi_hv##w##_8_mmi(uint8_t *_dst, \ · 12 unmodified lines
631	631	"psraw %[ftmp6], %[ftmp6], %[shift] \n\t" \
632	632	"packsswh %[ftmp5], %[ftmp5], %[ftmp6] \n\t" \
633	633	"pcmpgth %[ftmp7], %[ftmp5], %[ftmp7] \n\t" \
634		− "and %[ftmp3], %[ftmp5], %[ftmp7] \n\t" \
	634	+ "pand %[ftmp3], %[ftmp5], %[ftmp7] \n\t" \
635	635	"packushb %[ftmp3], %[ftmp3], %[ftmp3] \n\t" \
636	636	"gsswlc1 %[ftmp3], 0x03(%[dst]) \n\t" \
637	637	"gsswrc1 %[ftmp3], 0x00(%[dst]) \n\t" \
@@ -706,7 +706,7 @@ void ff_hevc_put_hevc_epel_bi_hv##w##_8_mmi(uint8_t *_dst, \ · 68 unmodified lines
706	706	"dmtc1 %[rtmp0], %[ftmp0] \n\t" \
707	707	"punpcklbh %[ftmp1], %[ftmp0], %[ftmp1] \n\t" \
708	708	"psrah %[ftmp1], %[ftmp1], %[ftmp0] \n\t" \
709		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
	709	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
710	710	\
711	711	"1: \n\t" \
712	712	"2: \n\t" \
@@ -771,7 +771,7 @@ void ff_hevc_put_hevc_epel_bi_hv##w##_8_mmi(uint8_t *_dst, \ · 58 unmodified lines
771	771	"li %[rtmp0], 0x06 \n\t" \
772	772	"dmtc1 %[rtmp0], %[ftmp0] \n\t" \
773	773	"punpcklwd %[offset], %[offset], %[offset] \n\t" \
774		− "xor %[ftmp2], %[ftmp2], %[ftmp2] \n\t" \
	774	+ "pxor %[ftmp2], %[ftmp2], %[ftmp2] \n\t" \
775	775	\
776	776	"1: \n\t" \
777	777	"li %[x], " #x_step " \n\t" \
@@ -821,7 +821,7 @@ void ff_hevc_put_hevc_epel_bi_hv##w##_8_mmi(uint8_t *_dst, \ · 43 unmodified lines
821	821	"psraw %[ftmp6], %[ftmp6], %[shift] \n\t" \
822	822	"packsswh %[ftmp5], %[ftmp5], %[ftmp6] \n\t" \
823	823	"pcmpgth %[ftmp7], %[ftmp5], %[ftmp2] \n\t" \
824		− "and %[ftmp3], %[ftmp5], %[ftmp7] \n\t" \
	824	+ "pand %[ftmp3], %[ftmp5], %[ftmp7] \n\t" \
825	825	"packushb %[ftmp3], %[ftmp3], %[ftmp3] \n\t" \
826	826	"gsswlc1 %[ftmp3], 0x03(%[dst]) \n\t" \
827	827	"gsswrc1 %[ftmp3], 0x00(%[dst]) \n\t" \
@@ -882,7 +882,7 @@ void ff_hevc_put_hevc_pel_bi_pixels##w##_8_mmi(uint8_t *_dst, \ · 54 unmodified lines
882	882	y = height; \
883	883	x = width >> 3; \
884	884	__asm__ volatile( \
885		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
	885	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
886	886	"li %[rtmp0], 0x06 \n\t" \
887	887	"dmtc1 %[rtmp0], %[ftmp1] \n\t" \
888	888	"li %[rtmp0], 0x10 \n\t" \
@@ -930,8 +930,8 @@ void ff_hevc_put_hevc_pel_bi_pixels##w##_8_mmi(uint8_t *_dst, \ · 41 unmodified lines
930	930	"packsswh %[ftmp4], %[ftmp4], %[ftmp5] \n\t" \
931	931	"pcmpgth %[ftmp3], %[ftmp2], %[ftmp0] \n\t" \
932	932	"pcmpgth %[ftmp5], %[ftmp4], %[ftmp0] \n\t" \
933		− "and %[ftmp2], %[ftmp2], %[ftmp3] \n\t" \
934		− "and %[ftmp4], %[ftmp4], %[ftmp5] \n\t" \
	933	+ "pand %[ftmp2], %[ftmp2], %[ftmp3] \n\t" \
	934	+ "pand %[ftmp4], %[ftmp4], %[ftmp5] \n\t" \
935	935	"packushb %[ftmp2], %[ftmp2], %[ftmp4] \n\t" \
936	936	"gssdlc1 %[ftmp2], 0x07(%[dst]) \n\t" \
937	937	"gssdrc1 %[ftmp2], 0x00(%[dst]) \n\t" \
@@ -1006,7 +1006,7 @@ void ff_hevc_put_hevc_qpel_uni_hv##w##_8_mmi(uint8_t *_dst, \ · 68 unmodified lines
1006	1006	"punpcklbh %[ftmp1], %[ftmp0], %[ftmp1] \n\t" \
1007	1007	"psrah %[ftmp1], %[ftmp1], %[ftmp0] \n\t" \
1008	1008	"psrah %[ftmp2], %[ftmp2], %[ftmp0] \n\t" \
1009		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
	1009	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
1010	1010	\
1011	1011	"1: \n\t" \
1012	1012	"2: \n\t" \
@@ -1139,9 +1139,9 @@ void ff_hevc_put_hevc_qpel_uni_hv##w##_8_mmi(uint8_t *_dst, \ · 126 unmodified lines
1139	1139	"packsswh %[ftmp3], %[ftmp3], %[ftmp5] \n\t" \
1140	1140	"paddh %[ftmp3], %[ftmp3], %[offset] \n\t" \
1141	1141	"psrah %[ftmp3], %[ftmp3], %[shift] \n\t" \
1142		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t" \
	1142	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t" \
1143	1143	"pcmpgth %[ftmp7], %[ftmp3], %[ftmp7] \n\t" \
1144		− "and %[ftmp3], %[ftmp3], %[ftmp7] \n\t" \
	1144	+ "pand %[ftmp3], %[ftmp3], %[ftmp7] \n\t" \
1145	1145	"packushb %[ftmp3], %[ftmp3], %[ftmp3] \n\t" \
1146	1146	"gsswlc1 %[ftmp3], 0x03(%[dst]) \n\t" \
1147	1147	"gsswrc1 %[ftmp3], 0x00(%[dst]) \n\t" \

libavcodec/mips/hpeldsp_mmi.c+13−13

@@ -676,14 +676,14 @@ inline void ff_put_no_rnd_pixels8_l2_8_mmi(uint8_t dst, const uint8_t src1, · 675 unmodified lines
676	676	PTR_ADDU "%[addr1], %[src2], %[src_stride2] \n\t"
677	677	MMI_ULDC1(%[ftmp3], %[addr1], 0x00)
678	678	PTR_ADDU "%[src1], %[src1], %[addr2] \n\t"
679		− "xor %[ftmp0], %[ftmp0], %[ftmp4] \n\t"
680		− "xor %[ftmp1], %[ftmp1], %[ftmp4] \n\t"
681		− "xor %[ftmp2], %[ftmp2], %[ftmp4] \n\t"
682		− "xor %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
	679	+ "pxor %[ftmp0], %[ftmp0], %[ftmp4] \n\t"
	680	+ "pxor %[ftmp1], %[ftmp1], %[ftmp4] \n\t"
	681	+ "pxor %[ftmp2], %[ftmp2], %[ftmp4] \n\t"
	682	+ "pxor %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
683	683	"pavgb %[ftmp0], %[ftmp0], %[ftmp2] \n\t"
684	684	"pavgb %[ftmp1], %[ftmp1], %[ftmp3] \n\t"
685		− "xor %[ftmp0], %[ftmp0], %[ftmp4] \n\t"
686		− "xor %[ftmp1], %[ftmp1], %[ftmp4] \n\t"
	685	+ "pxor %[ftmp0], %[ftmp0], %[ftmp4] \n\t"
	686	+ "pxor %[ftmp1], %[ftmp1], %[ftmp4] \n\t"
687	687	MMI_SDC1(%[ftmp0], %[dst], 0x00)
688	688	MMI_SDXC1(%[ftmp1], %[dst], %[dst_stride], 0x00)
689	689	PTR_ADDU "%[src2], %[src2], %[addr3] \n\t"
@@ -696,14 +696,14 @@ inline void ff_put_no_rnd_pixels8_l2_8_mmi(uint8_t dst, const uint8_t src1, · 6 unmodified lines
696	696	PTR_ADDU "%[addr1], %[src2], %[src_stride2] \n\t"
697	697	MMI_ULDC1(%[ftmp3], %[addr1], 0x00)
698	698	PTR_ADDU "%[src1], %[src1], %[addr2] \n\t"
699		− "xor %[ftmp0], %[ftmp0], %[ftmp4] \n\t"
700		− "xor %[ftmp1], %[ftmp1], %[ftmp4] \n\t"
701		− "xor %[ftmp2], %[ftmp2], %[ftmp4] \n\t"
702		− "xor %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
	699	+ "pxor %[ftmp0], %[ftmp0], %[ftmp4] \n\t"
	700	+ "pxor %[ftmp1], %[ftmp1], %[ftmp4] \n\t"
	701	+ "pxor %[ftmp2], %[ftmp2], %[ftmp4] \n\t"
	702	+ "pxor %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
703	703	"pavgb %[ftmp0], %[ftmp0], %[ftmp2] \n\t"
704	704	"pavgb %[ftmp1], %[ftmp1], %[ftmp3] \n\t"
705		− "xor %[ftmp0], %[ftmp0], %[ftmp4] \n\t"
706		− "xor %[ftmp1], %[ftmp1], %[ftmp4] \n\t"
	705	+ "pxor %[ftmp0], %[ftmp0], %[ftmp4] \n\t"
	706	+ "pxor %[ftmp1], %[ftmp1], %[ftmp4] \n\t"
707	707	MMI_SDC1(%[ftmp0], %[dst], 0x00)
708	708	MMI_SDXC1(%[ftmp1], %[dst], %[dst_stride], 0x00)
709	709	PTR_ADDU "%[src2], %[src2], %[addr3] \n\t"
@@ -846,7 +846,7 @@ void ff_put_pixels8_xy2_8_mmi(uint8_t block, const uint8_t pixels, · 136 unmodified lines
846	846	DECLARE_VAR_ADDRT;
847	847
848	848	__asm__ volatile (
849		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
	849	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
850	850	"dli %[addr0], 0x0f \n\t"
851	851	"pcmpeqw %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
852	852	"dmtc1 %[addr0], %[ftmp8] \n\t"

libavcodec/mips/idctdsp_mmi.c+1−1

@@ -154,7 +154,7 @@ void ff_add_pixels_clamped_mmi(const int16_t *block, · 153 unmodified lines
154	154	uint64_t tmp[1];
155	155	__asm__ volatile (
156	156	"li %[tmp0], 0x04 \n\t"
157		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	157	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
158	158	"1: \n\t"
159	159	MMI_LDC1(%[ftmp5], %[pixels], 0x00)
160	160	PTR_ADDU "%[pixels], %[pixels], %[line_size] \n\t"

libavcodec/mips/mpegvideo_mmi.c+47−47

@@ -53,13 +53,13 @@ void ff_dct_unquantize_h263_intra_mmi(MpegEncContext s, int16_t block, · 52 unmodified lines
53	53	nCoeffs = s->inter_scantable.raster_end[s->block_last_index[n]];
54	54
55	55	__asm__ volatile (
56		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	56	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
57	57	"packsswh %[qmul], %[qmul], %[qmul] \n\t"
58	58	"packsswh %[qmul], %[qmul], %[qmul] \n\t"
59	59	"packsswh %[qadd], %[qadd], %[qadd] \n\t"
60	60	"packsswh %[qadd], %[qadd], %[qadd] \n\t"
61	61	"psubh %[ftmp0], %[ftmp0], %[qadd] \n\t"
62		− "xor %[ftmp5], %[ftmp5], %[ftmp5] \n\t"
	62	+ "pxor %[ftmp5], %[ftmp5], %[ftmp5] \n\t"
63	63	".p2align 4 \n\t"
64	64
65	65	"1: \n\t"
@@ -72,12 +72,12 @@ void ff_dct_unquantize_h263_intra_mmi(MpegEncContext s, int16_t block, · 6 unmodified lines
72	72	"pmullh %[ftmp2], %[ftmp2], %[qmul] \n\t"
73	73	"pcmpgth %[ftmp3], %[ftmp3], %[ftmp5] \n\t"
74	74	"pcmpgth %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
75		− "xor %[ftmp1], %[ftmp1], %[ftmp3] \n\t"
76		− "xor %[ftmp2], %[ftmp2], %[ftmp4] \n\t"
	75	+ "pxor %[ftmp1], %[ftmp1], %[ftmp3] \n\t"
	76	+ "pxor %[ftmp2], %[ftmp2], %[ftmp4] \n\t"
77	77	"paddh %[ftmp1], %[ftmp1], %[ftmp0] \n\t"
78	78	"paddh %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
79		− "xor %[ftmp3], %[ftmp3], %[ftmp1] \n\t"
80		− "xor %[ftmp4], %[ftmp4], %[ftmp2] \n\t"
	79	+ "pxor %[ftmp3], %[ftmp3], %[ftmp1] \n\t"
	80	+ "pxor %[ftmp4], %[ftmp4], %[ftmp2] \n\t"
81	81	"pcmpeqh %[ftmp1], %[ftmp1], %[ftmp0] \n\t"
82	82	"pcmpeqh %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
83	83	"pandn %[ftmp1], %[ftmp1], %[ftmp3] \n\t"
@@ -116,11 +116,11 @@ void ff_dct_unquantize_h263_inter_mmi(MpegEncContext s, int16_t block, · 32 unmodified lines
116	116	__asm__ volatile (
117	117	"packsswh %[qmul], %[qmul], %[qmul] \n\t"
118	118	"packsswh %[qmul], %[qmul], %[qmul] \n\t"
119		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	119	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
120	120	"packsswh %[qadd], %[qadd], %[qadd] \n\t"
121	121	"packsswh %[qadd], %[qadd], %[qadd] \n\t"
122	122	"psubh %[ftmp0], %[ftmp0], %[qadd] \n\t"
123		− "xor %[ftmp5], %[ftmp5], %[ftmp5] \n\t"
	123	+ "pxor %[ftmp5], %[ftmp5], %[ftmp5] \n\t"
124	124	".p2align 4 \n\t"
125	125	"1: \n\t"
126	126	PTR_ADDU "%[addr0], %[block], %[nCoeffs] \n\t"
@@ -132,12 +132,12 @@ void ff_dct_unquantize_h263_inter_mmi(MpegEncContext s, int16_t block, · 5 unmodified lines
132	132	"pmullh %[ftmp2], %[ftmp2], %[qmul] \n\t"
133	133	"pcmpgth %[ftmp3], %[ftmp3], %[ftmp5] \n\t"
134	134	"pcmpgth %[ftmp4], %[ftmp4], %[ftmp5] \n\t"
135		− "xor %[ftmp1], %[ftmp1], %[ftmp3] \n\t"
136		− "xor %[ftmp2], %[ftmp2], %[ftmp4] \n\t"
	135	+ "pxor %[ftmp1], %[ftmp1], %[ftmp3] \n\t"
	136	+ "pxor %[ftmp2], %[ftmp2], %[ftmp4] \n\t"
137	137	"paddh %[ftmp1], %[ftmp1], %[ftmp0] \n\t"
138	138	"paddh %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
139		− "xor %[ftmp3], %[ftmp3], %[ftmp1] \n\t"
140		− "xor %[ftmp4], %[ftmp4], %[ftmp2] \n\t"
	139	+ "pxor %[ftmp3], %[ftmp3], %[ftmp1] \n\t"
	140	+ "pxor %[ftmp4], %[ftmp4], %[ftmp2] \n\t"
141	141	"pcmpeqh %[ftmp1], %[ftmp1], %[ftmp0] \n\t"
142	142	"pcmpeqh %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
143	143	"pandn %[ftmp1], %[ftmp1], %[ftmp3] \n\t"
@@ -201,18 +201,18 @@ void ff_dct_unquantize_mpeg1_intra_mmi(MpegEncContext s, int16_t block, · 57 unmodified lines
201	201	MMI_LDXC1(%[ftmp7], %[addr0], %[quant], 0x08)
202	202	"pmullh %[ftmp6], %[ftmp6], %[ftmp1] \n\t"
203	203	"pmullh %[ftmp7], %[ftmp7], %[ftmp1] \n\t"
204		− "xor %[ftmp8], %[ftmp8], %[ftmp8] \n\t"
205		− "xor %[ftmp9], %[ftmp9], %[ftmp9] \n\t"
	204	+ "pxor %[ftmp8], %[ftmp8], %[ftmp8] \n\t"
	205	+ "pxor %[ftmp9], %[ftmp9], %[ftmp9] \n\t"
206	206	"pcmpgth %[ftmp8], %[ftmp8], %[ftmp2] \n\t"
207	207	"pcmpgth %[ftmp9], %[ftmp9], %[ftmp3] \n\t"
208		− "xor %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
209		− "xor %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
	208	+ "pxor %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
	209	+ "pxor %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
210	210	"psubh %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
211	211	"psubh %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
212	212	"pmullh %[ftmp2], %[ftmp2], %[ftmp6] \n\t"
213	213	"pmullh %[ftmp3], %[ftmp3], %[ftmp7] \n\t"
214		− "xor %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
215		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
	214	+ "pxor %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
	215	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
216	216	"pcmpeqh %[ftmp6], %[ftmp6], %[ftmp4] \n\t"
217	217	"dli %[tmp0], 0x03 \n\t"
218	218	"pcmpeqh %[ftmp7], %[ftmp7], %[ftmp5] \n\t"
@@ -221,10 +221,10 @@ void ff_dct_unquantize_mpeg1_intra_mmi(MpegEncContext s, int16_t block, · 2 unmodified lines
221	221	"psrah %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
222	222	"psubh %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
223	223	"psubh %[ftmp3], %[ftmp3], %[ftmp0] \n\t"
224		− "or %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
225		− "or %[ftmp3], %[ftmp3], %[ftmp0] \n\t"
226		− "xor %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
227		− "xor %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
	224	+ "por %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
	225	+ "por %[ftmp3], %[ftmp3], %[ftmp0] \n\t"
	226	+ "pxor %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
	227	+ "pxor %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
228	228	"psubh %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
229	229	"psubh %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
230	230	"pandn %[ftmp6], %[ftmp6], %[ftmp2] \n\t"
@@ -287,12 +287,12 @@ void ff_dct_unquantize_mpeg1_inter_mmi(MpegEncContext s, int16_t block, · 56 unmodified lines
287	287	MMI_LDXC1(%[ftmp7], %[addr0], %[quant], 0x08)
288	288	"pmullh %[ftmp6], %[ftmp6], %[ftmp1] \n\t"
289	289	"pmullh %[ftmp7], %[ftmp7], %[ftmp1] \n\t"
290		− "xor %[ftmp8], %[ftmp8], %[ftmp8] \n\t"
291		− "xor %[ftmp9], %[ftmp9], %[ftmp9] \n\t"
	290	+ "pxor %[ftmp8], %[ftmp8], %[ftmp8] \n\t"
	291	+ "pxor %[ftmp9], %[ftmp9], %[ftmp9] \n\t"
292	292	"pcmpgth %[ftmp8], %[ftmp8], %[ftmp2] \n\t"
293	293	"pcmpgth %[ftmp9], %[ftmp9], %[ftmp3] \n\t"
294		− "xor %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
295		− "xor %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
	294	+ "pxor %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
	295	+ "pxor %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
296	296	"psubh %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
297	297	"psubh %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
298	298	"paddh %[ftmp2], %[ftmp2], %[ftmp2] \n\t"
@@ -301,8 +301,8 @@ void ff_dct_unquantize_mpeg1_inter_mmi(MpegEncContext s, int16_t block, · 2 unmodified lines
301	301	"paddh %[ftmp3], %[ftmp3], %[ftmp0] \n\t"
302	302	"pmullh %[ftmp2], %[ftmp2], %[ftmp6] \n\t"
303	303	"pmullh %[ftmp3], %[ftmp3], %[ftmp7] \n\t"
304		− "xor %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
305		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
	304	+ "pxor %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
	305	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
306	306	"pcmpeqh %[ftmp6], %[ftmp6], %[ftmp4] \n\t"
307	307	"dli %[tmp0], 0x04 \n\t"
308	308	"pcmpeqh %[ftmp7], %[ftmp7], %[ftmp5] \n\t"
@@ -311,10 +311,10 @@ void ff_dct_unquantize_mpeg1_inter_mmi(MpegEncContext s, int16_t block, · 2 unmodified lines
311	311	"psrah %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
312	312	"psubh %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
313	313	"psubh %[ftmp3], %[ftmp3], %[ftmp0] \n\t"
314		− "or %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
315		− "or %[ftmp3], %[ftmp3], %[ftmp0] \n\t"
316		− "xor %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
317		− "xor %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
	314	+ "por %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
	315	+ "por %[ftmp3], %[ftmp3], %[ftmp0] \n\t"
	316	+ "pxor %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
	317	+ "pxor %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
318	318	"psubh %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
319	319	"psubh %[ftmp3], %[ftmp3], %[ftmp9] \n\t"
320	320	"pandn %[ftmp6], %[ftmp6], %[ftmp2] \n\t"
@@ -386,26 +386,26 @@ void ff_dct_unquantize_mpeg2_intra_mmi(MpegEncContext s, int16_t block, · 65 unmodified lines
386	386	MMI_LDXC1(%[ftmp6], %[addr0], %[quant], 0x08)
387	387	"pmullh %[ftmp5], %[ftmp5], %[ftmp9] \n\t"
388	388	"pmullh %[ftmp6], %[ftmp6], %[ftmp9] \n\t"
389		− "xor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
390		− "xor %[ftmp8], %[ftmp8], %[ftmp8] \n\t"
	389	+ "pxor %[ftmp7], %[ftmp7], %[ftmp7] \n\t"
	390	+ "pxor %[ftmp8], %[ftmp8], %[ftmp8] \n\t"
391	391	"pcmpgth %[ftmp7], %[ftmp7], %[ftmp1] \n\t"
392	392	"pcmpgth %[ftmp8], %[ftmp8], %[ftmp2] \n\t"
393		− "xor %[ftmp1], %[ftmp1], %[ftmp7] \n\t"
394		− "xor %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
	393	+ "pxor %[ftmp1], %[ftmp1], %[ftmp7] \n\t"
	394	+ "pxor %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
395	395	"psubh %[ftmp1], %[ftmp1], %[ftmp7] \n\t"
396	396	"psubh %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
397	397	"pmullh %[ftmp1], %[ftmp1], %[ftmp5] \n\t"
398	398	"pmullh %[ftmp2], %[ftmp2], %[ftmp6] \n\t"
399		− "xor %[ftmp5], %[ftmp5], %[ftmp5] \n\t"
400		− "xor %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
	399	+ "pxor %[ftmp5], %[ftmp5], %[ftmp5] \n\t"
	400	+ "pxor %[ftmp6], %[ftmp6], %[ftmp6] \n\t"
401	401	"pcmpeqh %[ftmp5], %[ftmp5], %[ftmp3] \n\t"
402	402	"dli %[tmp0], 0x03 \n\t"
403	403	"pcmpeqh %[ftmp6] , %[ftmp6], %[ftmp4] \n\t"
404	404	"mtc1 %[tmp0], %[ftmp3] \n\t"
405	405	"psrah %[ftmp1], %[ftmp1], %[ftmp3] \n\t"
406	406	"psrah %[ftmp2], %[ftmp2], %[ftmp3] \n\t"
407		− "xor %[ftmp1], %[ftmp1], %[ftmp7] \n\t"
408		− "xor %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
	407	+ "pxor %[ftmp1], %[ftmp1], %[ftmp7] \n\t"
	408	+ "pxor %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
409	409	"psubh %[ftmp1], %[ftmp1], %[ftmp7] \n\t"
410	410	"psubh %[ftmp2], %[ftmp2], %[ftmp8] \n\t"
411	411	"pandn %[ftmp5], %[ftmp5], %[ftmp1] \n\t"
@@ -445,16 +445,16 @@ void ff_denoise_dct_mmi(MpegEncContext s, int16_t block) · 33 unmodified lines
445	445	s->dct_count[intra]++;
446	446
447	447	__asm__ volatile(
448		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	448	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
449	449	"1: \n\t"
450	450	MMI_LDC1(%[ftmp1], %[block], 0x00)
451		− "xor %[ftmp2], %[ftmp2], %[ftmp2] \n\t"
	451	+ "pxor %[ftmp2], %[ftmp2], %[ftmp2] \n\t"
452	452	MMI_LDC1(%[ftmp3], %[block], 0x08)
453		− "xor %[ftmp4], %[ftmp4], %[ftmp4] \n\t"
	453	+ "pxor %[ftmp4], %[ftmp4], %[ftmp4] \n\t"
454	454	"pcmpgth %[ftmp2], %[ftmp2], %[ftmp1] \n\t"
455	455	"pcmpgth %[ftmp4], %[ftmp4], %[ftmp3] \n\t"
456		− "xor %[ftmp1], %[ftmp1], %[ftmp2] \n\t"
457		− "xor %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
	456	+ "pxor %[ftmp1], %[ftmp1], %[ftmp2] \n\t"
	457	+ "pxor %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
458	458	"psubh %[ftmp1], %[ftmp1], %[ftmp2] \n\t"
459	459	"psubh %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
460	460	MMI_LDC1(%[ftmp6], %[offset], 0x00)
@@ -463,8 +463,8 @@ void ff_denoise_dct_mmi(MpegEncContext s, int16_t block) · 2 unmodified lines
463	463	MMI_LDC1(%[ftmp6], %[offset], 0x08)
464	464	"mov.d %[ftmp7], %[ftmp3] \n\t"
465	465	"psubush %[ftmp3], %[ftmp3], %[ftmp6] \n\t"
466		− "xor %[ftmp1], %[ftmp1], %[ftmp2] \n\t"
467		− "xor %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
	466	+ "pxor %[ftmp1], %[ftmp1], %[ftmp2] \n\t"
	467	+ "pxor %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
468	468	"psubh %[ftmp1], %[ftmp1], %[ftmp2] \n\t"
469	469	"psubh %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
470	470	MMI_SDC1(%[ftmp1], %[block], 0x00)

libavcodec/mips/pixblockdsp_mmi.c+4−4

@@ -33,7 +33,7 @@ void ff_get_pixels_8_mmi(int16_t av_restrict block, const uint8_t pixels, · 32 unmodified lines
33	33	DECLARE_VAR_ADDRT;
34	34
35	35	__asm__ volatile (
36		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	36	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
37	37
38	38	MMI_LDC1(%[ftmp1], %[pixels], 0x00)
39	39	MMI_LDXC1(%[ftmp2], %[pixels], %[stride], 0x00)
@@ -103,12 +103,12 @@ void ff_diff_pixels_mmi(int16_t av_restrict block, const uint8_t src1, · 63 unmodified lines
103	103
104	104	__asm__ volatile (
105	105	"li %[tmp0], 0x08 \n\t"
106		− "xor %[ftmp4], %[ftmp4], %[ftmp4] \n\t"
	106	+ "pxor %[ftmp4], %[ftmp4], %[ftmp4] \n\t"
107	107	"1: \n\t"
108	108	MMI_LDC1(%[ftmp0], %[src1], 0x00)
109		− "or %[ftmp1], %[ftmp0], %[ftmp0] \n\t"
	109	+ "por %[ftmp1], %[ftmp0], %[ftmp0] \n\t"
110	110	MMI_LDC1(%[ftmp2], %[src2], 0x00)
111		− "or %[ftmp3], %[ftmp2], %[ftmp2] \n\t"
	111	+ "por %[ftmp3], %[ftmp2], %[ftmp2] \n\t"
112	112	"punpcklbh %[ftmp0], %[ftmp0], %[ftmp4] \n\t"
113	113	"punpckhbh %[ftmp1], %[ftmp1], %[ftmp4] \n\t"
114	114	"punpcklbh %[ftmp2], %[ftmp2], %[ftmp4] \n\t"

libavcodec/mips/simple_idct_mmi.c+7−7

@@ -133,7 +133,7 @@ void ff_simple_idct_8_mmi(int16_t *block) · 132 unmodified lines
133	133	"psllh $f28, "#src1", $f30 \n\t" \
134	134	"dmtc1 $9, $f31 \n\t" \
135	135	"punpcklhw $f29, $f28, $f28 \n\t" \
136		− "and $f29, $f29, $f31 \n\t" \
	136	+ "pand $f29, $f29, $f31 \n\t" \
137	137	"paddw $f28, $f28, $f29 \n\t" \
138	138	"punpcklwd "#src1", $f28, $f28 \n\t" \
139	139	"punpcklwd "#src2", $f28, $f28 \n\t" \
@@ -268,9 +268,9 @@ void ff_simple_idct_8_mmi(int16_t *block) · 128 unmodified lines
268	268	"punpcklwd $f8, $f27, $f29 \n\t"
269	269	"punpckhwd $f12, $f27, $f29 \n\t"
270	270
271		− "or $f26, $f2, $f6 \n\t"
272		− "or $f26, $f26, $f10 \n\t"
273		− "or $f26, $f26, $f14 \n\t"
	271	+ "por $f26, $f2, $f6 \n\t"
	272	+ "por $f26, $f26, $f10 \n\t"
	273	+ "por $f26, $f26, $f14 \n\t"
274	274	"dmfc1 $10, $f26 \n\t"
275	275	"bnez $10, 1f \n\t"
276	276	/* case1: In this case, row[1,3,5,7] are all zero */
@@ -338,9 +338,9 @@ void ff_simple_idct_8_mmi(int16_t *block) · 61 unmodified lines
338	338	"punpcklwd $f9, $f27, $f29 \n\t"
339	339	"punpckhwd $f13, $f27, $f29 \n\t"
340	340
341		− "or $f26, $f3, $f7 \n\t"
342		− "or $f26, $f26, $f11 \n\t"
343		− "or $f26, $f26, $f15 \n\t"
	341	+ "por $f26, $f3, $f7 \n\t"
	342	+ "por $f26, $f26, $f11 \n\t"
	343	+ "por $f26, $f26, $f15 \n\t"
344	344	"dmfc1 $10, $f26 \n\t"
345	345	"bnez $10, 1f \n\t"
346	346	/* case1: In this case, row[1,3,5,7] are all zero */

libavcodec/mips/vc1dsp_mmi.c+17−17

@@ -134,7 +134,7 @@ void ff_vc1_inv_trans_8x8_dc_mmi(uint8_t dest, ptrdiff_t linesize, int16_t blo · 133 unmodified lines
134	134	dc = (3 * dc + 16) >> 5;
135	135
136	136	__asm__ volatile(
137		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	137	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
138	138	"pshufh %[dc], %[dc], %[ftmp0] \n\t"
139	139	"li %[count], 0x02 \n\t"
140	140
@@ -425,7 +425,7 @@ void ff_vc1_inv_trans_8x4_dc_mmi(uint8_t dest, ptrdiff_t linesize, int16_t blo · 284 unmodified lines
425	425	dc = (17 * dc + 64) >> 7;
426	426
427	427	__asm__ volatile(
428		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	428	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
429	429	"pshufh %[dc], %[dc], %[ftmp0] \n\t"
430	430
431	431	MMI_LDC1(%[ftmp1], %[dest0], 0x00)
@@ -705,7 +705,7 @@ void ff_vc1_inv_trans_8x4_mmi(uint8_t dest, ptrdiff_t linesize, int16_t block) · 273 unmodified lines
705	705	MMI_LWC1(%[ftmp3], %[tmp0], 0x00)
706	706	PTR_ADDU "%[tmp0], %[tmp0], %[linesize] \n\t"
707	707	MMI_LWC1(%[ftmp4], %[tmp0], 0x00)
708		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	708	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
709	709	"punpcklbh %[ftmp1], %[ftmp1], %[ftmp0] \n\t"
710	710	"punpcklbh %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
711	711	"punpcklbh %[ftmp3], %[ftmp3], %[ftmp0] \n\t"
@@ -829,7 +829,7 @@ void ff_vc1_inv_trans_8x4_mmi(uint8_t dest, ptrdiff_t linesize, int16_t block) · 117 unmodified lines
829	829	MMI_LWC1(%[ftmp3], %[tmp0], 0x04)
830	830	PTR_ADDU "%[tmp0], %[tmp0], %[linesize] \n\t"
831	831	MMI_LWC1(%[ftmp4], %[tmp0], 0x04)
832		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	832	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
833	833	"punpcklbh %[ftmp1], %[ftmp1], %[ftmp0] \n\t"
834	834	"punpcklbh %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
835	835	"punpcklbh %[ftmp3], %[ftmp3], %[ftmp0] \n\t"
@@ -877,7 +877,7 @@ void ff_vc1_inv_trans_4x8_dc_mmi(uint8_t dest, ptrdiff_t linesize, int16_t blo · 41 unmodified lines
877	877	dc = (12 * dc + 64) >> 7;
878	878
879	879	__asm__ volatile(
880		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	880	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
881	881	"pshufh %[dc], %[dc], %[ftmp0] \n\t"
882	882
883	883	MMI_LWC1(%[ftmp1], %[dest0], 0x00)
@@ -1058,7 +1058,7 @@ void ff_vc1_inv_trans_4x8_mmi(uint8_t dest, ptrdiff_t linesize, int16_t block) · 174 unmodified lines
1058	1058	MMI_LWC1(%[ftmp7], %[tmp0], 0x00)
1059	1059	PTR_ADDU "%[tmp0], %[tmp0], %[linesize] \n\t"
1060	1060	MMI_LWC1(%[ftmp8], %[tmp0], 0x00)
1061		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1061	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1062	1062	"punpcklbh %[ftmp1], %[ftmp1], %[ftmp0] \n\t"
1063	1063	"punpcklbh %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
1064	1064	"punpcklbh %[ftmp3], %[ftmp3], %[ftmp0] \n\t"
@@ -1133,7 +1133,7 @@ void ff_vc1_inv_trans_4x4_dc_mmi(uint8_t dest, ptrdiff_t linesize, int16_t blo · 68 unmodified lines
1133	1133	dc = (17 * dc + 64) >> 7;
1134	1134
1135	1135	__asm__ volatile(
1136		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1136	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1137	1137	"pshufh %[dc], %[dc], %[ftmp0] \n\t"
1138	1138
1139	1139	MMI_LWC1(%[ftmp1], %[dest0], 0x00)
@@ -1339,7 +1339,7 @@ void ff_vc1_inv_trans_4x4_mmi(uint8_t dest, ptrdiff_t linesize, int16_t block) · 199 unmodified lines
1339	1339	MMI_LWC1(%[ftmp3], %[tmp0], 0x00)
1340	1340	PTR_ADDU "%[tmp0], %[tmp0], %[linesize] \n\t"
1341	1341	MMI_LWC1(%[ftmp4], %[tmp0], 0x00)
1342		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1342	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1343	1343	"punpcklbh %[ftmp1], %[ftmp1], %[ftmp0] \n\t"
1344	1344	"punpcklbh %[ftmp2], %[ftmp2], %[ftmp0] \n\t"
1345	1345	"punpcklbh %[ftmp3], %[ftmp3], %[ftmp0] \n\t"
@@ -1664,7 +1664,7 @@ static void vc1_put_ver_16b_shift2_mmi(int16_t *dst, · 318 unmodified lines
1664	1664	DECLARE_VAR_ADDRT;
1665	1665
1666	1666	__asm__ volatile(
1667		− "xor $f0, $f0, $f0 \n\t"
	1667	+ "pxor $f0, $f0, $f0 \n\t"
1668	1668	"li $8, 0x03 \n\t"
1669	1669	LOAD_ROUNDER_MMI("%[rnd]")
1670	1670	"ldc1 $f12, %[ff_pw_9] \n\t"
@@ -1771,7 +1771,7 @@ static void OPNAME ## vc1_shift2_mmi(uint8_t dst, const uint8_t src, \ · 100 unmodified lines
1771	1771	rnd = 8 - rnd; \
1772	1772	\
1773	1773	__asm__ volatile( \
1774		− "xor $f0, $f0, $f0 \n\t" \
	1774	+ "pxor $f0, $f0, $f0 \n\t" \
1775	1775	"li $10, 0x08 \n\t" \
1776	1776	LOAD_ROUNDER_MMI("%[rnd]") \
1777	1777	"ldc1 $f12, %[ff_pw_9] \n\t" \
@@ -1898,7 +1898,7 @@ vc1_put_ver_16b_ ## NAME ## _mmi(int16_t dst, const uint8_t src, \ · 120 unmodified lines
1898	1898	src -= src_stride; \
1899	1899	\
1900	1900	__asm__ volatile( \
1901		− "xor $f0, $f0, $f0 \n\t" \
	1901	+ "pxor $f0, $f0, $f0 \n\t" \
1902	1902	LOAD_ROUNDER_MMI("%[rnd]") \
1903	1903	"ldc1 $f10, %[ff_pw_53] \n\t" \
1904	1904	"ldc1 $f12, %[ff_pw_18] \n\t" \
@@ -1973,7 +1973,7 @@ OPNAME ## vc1_hor_16b_ ## NAME ## _mmi(uint8_t *dst, mips_reg stride, \ · 68 unmodified lines
1973	1973	rnd -= (-4+58+13-3)256; / Add -256 bias */ \
1974	1974	\
1975	1975	__asm__ volatile( \
1976		− "xor $f0, $f0, $f0 \n\t" \
	1976	+ "pxor $f0, $f0, $f0 \n\t" \
1977	1977	LOAD_ROUNDER_MMI("%[rnd]") \
1978	1978	"ldc1 $f10, %[ff_pw_53] \n\t" \
1979	1979	"ldc1 $f12, %[ff_pw_18] \n\t" \
@@ -2023,7 +2023,7 @@ OPNAME ## vc1_## NAME ## _mmi(uint8_t dst, const uint8_t src, \ · 43 unmodified lines
2023	2023	rnd = 32-rnd; \
2024	2024	\
2025	2025	__asm__ volatile ( \
2026		− "xor $f0, $f0, $f0 \n\t" \
	2026	+ "pxor $f0, $f0, $f0 \n\t" \
2027	2027	LOAD_ROUNDER_MMI("%[rnd]") \
2028	2028	"ldc1 $f10, %[ff_pw_53] \n\t" \
2029	2029	"ldc1 $f12, %[ff_pw_18] \n\t" \
@@ -2259,7 +2259,7 @@ void ff_put_no_rnd_vc1_chroma_mc8_mmi(uint8_t dst / align 8 */, · 229 unmodified lines
2259	2259
2260	2260	__asm__ volatile(
2261	2261	"li %[tmp0], 0x06 \n\t"
2262		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	2262	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
2263	2263	"mtc1 %[tmp0], %[ftmp9] \n\t"
2264	2264	"pshufh %[A], %[A], %[ftmp0] \n\t"
2265	2265	"pshufh %[B], %[B], %[ftmp0] \n\t"
@@ -2314,7 +2314,7 @@ void ff_put_no_rnd_vc1_chroma_mc4_mmi(uint8_t dst / align 8 */, · 48 unmodified lines
2314	2314
2315	2315	__asm__ volatile(
2316	2316	"li %[tmp0], 0x06 \n\t"
2317		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	2317	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
2318	2318	"mtc1 %[tmp0], %[ftmp5] \n\t"
2319	2319	"pshufh %[A], %[A], %[ftmp0] \n\t"
2320	2320	"pshufh %[B], %[B], %[ftmp0] \n\t"
@@ -2367,7 +2367,7 @@ void ff_avg_no_rnd_vc1_chroma_mc8_mmi(uint8_t dst / align 8 */, · 46 unmodified lines
2367	2367
2368	2368	__asm__ volatile(
2369	2369	"li %[tmp0], 0x06 \n\t"
2370		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	2370	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
2371	2371	"mtc1 %[tmp0], %[ftmp9] \n\t"
2372	2372	"pshufh %[A], %[A], %[ftmp0] \n\t"
2373	2373	"pshufh %[B], %[B], %[ftmp0] \n\t"
@@ -2425,7 +2425,7 @@ void ff_avg_no_rnd_vc1_chroma_mc4_mmi(uint8_t dst / align 8 */, · 51 unmodified lines
2425	2425
2426	2426	__asm__ volatile(
2427	2427	"li %[tmp0], 0x06 \n\t"
2428		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	2428	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
2429	2429	"mtc1 %[tmp0], %[ftmp5] \n\t"
2430	2430	"pshufh %[A], %[A], %[ftmp0] \n\t"
2431	2431	"pshufh %[B], %[B], %[ftmp0] \n\t"

libavcodec/mips/vp3dsp_idct_mmi.c+66−66

@@ -34,7 +34,7 @@ static void idct_row_mmi(int16_t *input) · 33 unmodified lines
34	34	double ftmp[23];
35	35	uint64_t tmp[2];
36	36	__asm__ volatile (
37		− "xor %[ftmp10], %[ftmp10], %[ftmp10] \n\t"
	37	+ "pxor %[ftmp10], %[ftmp10], %[ftmp10] \n\t"
38	38	LOAD_CONST(%[csth_1], 1)
39	39	"li %[tmp0], 0x02 \n\t"
40	40	"1: \n\t"
@@ -51,14 +51,14 @@ static void idct_row_mmi(int16_t *input) · 10 unmodified lines
51	51	LOAD_CONST(%[ftmp9], 12785)
52	52	"pmulhh %[A], %[ftmp9], %[ftmp7] \n\t"
53	53	"pcmpgth %[C], %[ftmp10], %[ftmp1] \n\t"
54		− "or %[mask], %[C], %[csth_1] \n\t"
	54	+ "por %[mask], %[C], %[csth_1] \n\t"
55	55	"pmullh %[B], %[ftmp1], %[mask] \n\t"
56	56	"pmulhuh %[B], %[ftmp8], %[B] \n\t"
57	57	"pmullh %[B], %[B], %[mask] \n\t"
58	58	"paddh %[A], %[A], %[B] \n\t"
59	59	"paddh %[A], %[A], %[C] \n\t"
60	60	"pcmpgth %[D], %[ftmp10], %[ftmp7] \n\t"
61		− "or %[mask], %[D], %[csth_1] \n\t"
	61	+ "por %[mask], %[D], %[csth_1] \n\t"
62	62	"pmullh %[ftmp7], %[ftmp7], %[mask] \n\t"
63	63	"pmulhuh %[B], %[ftmp8], %[ftmp7] \n\t"
64	64	"pmullh %[B], %[B], %[mask] \n\t"
@@ -69,12 +69,12 @@ static void idct_row_mmi(int16_t *input) · 4 unmodified lines
69	69	LOAD_CONST(%[ftmp8], 54491)
70	70	LOAD_CONST(%[ftmp9], 36410)
71	71	"pcmpgth %[Ad], %[ftmp10], %[ftmp5] \n\t"
72		− "or %[mask], %[Ad], %[csth_1] \n\t"
	72	+ "por %[mask], %[Ad], %[csth_1] \n\t"
73	73	"pmullh %[ftmp1], %[ftmp5], %[mask] \n\t"
74	74	"pmulhuh %[C], %[ftmp9], %[ftmp1] \n\t"
75	75	"pmullh %[C], %[C], %[mask] \n\t"
76	76	"pcmpgth %[Bd], %[ftmp10], %[ftmp3] \n\t"
77		− "or %[mask], %[Bd], %[csth_1] \n\t"
	77	+ "por %[mask], %[Bd], %[csth_1] \n\t"
78	78	"pmullh %[D], %[ftmp3], %[mask] \n\t"
79	79	"pmulhuh %[D], %[ftmp8], %[D] \n\t"
80	80	"pmullh %[D], %[D], %[mask] \n\t"
@@ -82,12 +82,12 @@ static void idct_row_mmi(int16_t *input) · 1 unmodified line
82	82	"paddh %[C], %[C], %[Ad] \n\t"
83	83	"paddh %[C], %[C], %[Bd] \n\t"
84	84	"pcmpgth %[Bd], %[ftmp10], %[ftmp3] \n\t"
85		− "or %[mask], %[Bd], %[csth_1] \n\t"
	85	+ "por %[mask], %[Bd], %[csth_1] \n\t"
86	86	"pmullh %[ftmp1], %[ftmp3], %[mask] \n\t"
87	87	"pmulhuh %[D], %[ftmp9], %[ftmp1] \n\t"
88	88	"pmullh %[D], %[D], %[mask] \n\t"
89	89	"pcmpgth %[Ed], %[ftmp10], %[ftmp5] \n\t"
90		− "or %[mask], %[Ed], %[csth_1] \n\t"
	90	+ "por %[mask], %[Ed], %[csth_1] \n\t"
91	91	"pmullh %[Ad], %[ftmp5], %[mask] \n\t"
92	92	"pmulhuh %[Ad], %[ftmp8], %[Ad] \n\t"
93	93	"pmullh %[Ad], %[Ad], %[mask] \n\t"
@@ -98,14 +98,14 @@ static void idct_row_mmi(int16_t *input) · 4 unmodified lines
98	98	LOAD_CONST(%[ftmp8], 46341)
99	99	"psubh %[Ad], %[A], %[C] \n\t"
100	100	"pcmpgth %[Bd], %[ftmp10], %[Ad] \n\t"
101		− "or %[mask], %[Bd], %[csth_1] \n\t"
	101	+ "por %[mask], %[Bd], %[csth_1] \n\t"
102	102	"pmullh %[Ad], %[Ad], %[mask] \n\t"
103	103	"pmulhuh %[Ad], %[ftmp8], %[Ad] \n\t"
104	104	"pmullh %[Ad], %[Ad], %[mask] \n\t"
105	105	"paddh %[Ad], %[Ad], %[Bd] \n\t"
106	106	"psubh %[Bd], %[B], %[D] \n\t"
107	107	"pcmpgth %[Cd], %[ftmp10], %[Bd] \n\t"
108		− "or %[mask], %[Cd], %[csth_1] \n\t"
	108	+ "por %[mask], %[Cd], %[csth_1] \n\t"
109	109	"pmullh %[Bd], %[Bd], %[mask] \n\t"
110	110	"pmulhuh %[Bd], %[ftmp8], %[Bd] \n\t"
111	111	"pmullh %[Bd], %[Bd], %[mask] \n\t"
@@ -114,14 +114,14 @@ static void idct_row_mmi(int16_t *input) · 2 unmodified lines
114	114	"paddh %[Dd], %[B], %[D] \n\t"
115	115	"paddh %[A], %[ftmp0], %[ftmp4] \n\t"
116	116	"pcmpgth %[B], %[ftmp10], %[A] \n\t"
117		− "or %[mask], %[B], %[csth_1] \n\t"
	117	+ "por %[mask], %[B], %[csth_1] \n\t"
118	118	"pmullh %[A], %[A], %[mask] \n\t"
119	119	"pmulhuh %[A], %[ftmp8], %[A] \n\t"
120	120	"pmullh %[A], %[A], %[mask] \n\t"
121	121	"paddh %[A], %[A], %[B] \n\t"
122	122	"psubh %[B], %[ftmp0], %[ftmp4] \n\t"
123	123	"pcmpgth %[C], %[ftmp10], %[B] \n\t"
124		− "or %[mask], %[C], %[csth_1] \n\t"
	124	+ "por %[mask], %[C], %[csth_1] \n\t"
125	125	"pmullh %[B], %[B], %[mask] \n\t"
126	126	"pmulhuh %[B], %[ftmp8], %[B] \n\t"
127	127	"pmullh %[B], %[B], %[mask] \n\t"
@@ -131,14 +131,14 @@ static void idct_row_mmi(int16_t *input) · 3 unmodified lines
131	131	LOAD_CONST(%[ftmp9], 25080)
132	132	"pmulhh %[C], %[ftmp9], %[ftmp6] \n\t"
133	133	"pcmpgth %[D], %[ftmp10], %[ftmp2] \n\t"
134		− "or %[mask], %[D], %[csth_1] \n\t"
	134	+ "por %[mask], %[D], %[csth_1] \n\t"
135	135	"pmullh %[Ed], %[ftmp2], %[mask] \n\t"
136	136	"pmulhuh %[Ed], %[ftmp8], %[Ed] \n\t"
137	137	"pmullh %[Ed], %[Ed], %[mask] \n\t"
138	138	"paddh %[C], %[C], %[Ed] \n\t"
139	139	"paddh %[C], %[C], %[D] \n\t"
140	140	"pcmpgth %[Ed], %[ftmp10], %[ftmp6] \n\t"
141		− "or %[mask], %[Ed], %[csth_1] \n\t"
	141	+ "por %[mask], %[Ed], %[csth_1] \n\t"
142	142	"pmullh %[ftmp6], %[ftmp6], %[mask] \n\t"
143	143	"pmulhuh %[D], %[ftmp8], %[ftmp6] \n\t"
144	144	"pmullh %[D], %[D], %[mask] \n\t"
@@ -193,7 +193,7 @@ static void idct_column_true_mmi(uint8_t dst, int stride, int16_t input) · 48 unmodified lines
193	193	for (int i = 0; i < 8; ++i)
194	194	temp_value[i] = av_clip_uint8(128 + ((46341 * input[i << 3] + (8 << 16)) >> 20));
195	195	__asm__ volatile (
196		− "xor %[ftmp10], %[ftmp10], %[ftmp10] \n\t"
	196	+ "pxor %[ftmp10], %[ftmp10], %[ftmp10] \n\t"
197	197	"li %[tmp0], 0x02 \n\t"
198	198	"1: \n\t"
199	199	"ldc1 %[ftmp0], 0x00(%[input]) \n\t"
@@ -213,14 +213,14 @@ static void idct_column_true_mmi(uint8_t dst, int stride, int16_t input) · 13 unmodified lines
213	213	LOAD_CONST(%[Gd], 1)
214	214	"pmulhh %[A], %[ftmp9], %[ftmp7] \n\t"
215	215	"pcmpgth %[C], %[ftmp10], %[ftmp1] \n\t"
216		− "or %[mask], %[C], %[Gd] \n\t"
	216	+ "por %[mask], %[C], %[Gd] \n\t"
217	217	"pmullh %[B], %[ftmp1], %[mask] \n\t"
218	218	"pmulhuh %[B], %[ftmp8], %[B] \n\t"
219	219	"pmullh %[B], %[B], %[mask] \n\t"
220	220	"paddh %[A], %[A], %[B] \n\t"
221	221	"paddh %[A], %[A], %[C] \n\t"
222	222	"pcmpgth %[D], %[ftmp10], %[ftmp7] \n\t"
223		− "or %[mask], %[D], %[Gd] \n\t"
	223	+ "por %[mask], %[D], %[Gd] \n\t"
224	224	"pmullh %[Ad], %[ftmp7], %[mask] \n\t"
225	225	"pmulhuh %[B], %[ftmp8], %[Ad] \n\t"
226	226	"pmullh %[B], %[B], %[mask] \n\t"
@@ -231,12 +231,12 @@ static void idct_column_true_mmi(uint8_t dst, int stride, int16_t input) · 4 unmodified lines
231	231	LOAD_CONST(%[ftmp8], 54491)
232	232	LOAD_CONST(%[ftmp9], 36410)
233	233	"pcmpgth %[Ad], %[ftmp10], %[ftmp5] \n\t"
234		− "or %[mask], %[Ad], %[Gd] \n\t"
	234	+ "por %[mask], %[Ad], %[Gd] \n\t"
235	235	"pmullh %[Cd], %[ftmp5], %[mask] \n\t"
236	236	"pmulhuh %[C], %[ftmp9], %[Cd] \n\t"
237	237	"pmullh %[C], %[C], %[mask] \n\t"
238	238	"pcmpgth %[Bd], %[ftmp10], %[ftmp3] \n\t"
239		− "or %[mask], %[Bd], %[Gd] \n\t"
	239	+ "por %[mask], %[Bd], %[Gd] \n\t"
240	240	"pmullh %[D], %[ftmp3], %[mask] \n\t"
241	241	"pmulhuh %[D], %[ftmp8], %[D] \n\t"
242	242	"pmullh %[D], %[D], %[mask] \n\t"
@@ -244,12 +244,12 @@ static void idct_column_true_mmi(uint8_t dst, int stride, int16_t input) · 1 unmodified line
244	244	"paddh %[C], %[C], %[Ad] \n\t"
245	245	"paddh %[C], %[C], %[Bd] \n\t"
246	246	"pcmpgth %[Bd], %[ftmp10], %[ftmp3] \n\t"
247		− "or %[mask], %[Bd], %[Gd] \n\t"
	247	+ "por %[mask], %[Bd], %[Gd] \n\t"
248	248	"pmullh %[Cd], %[ftmp3], %[mask] \n\t"
249	249	"pmulhuh %[D], %[ftmp9], %[Cd] \n\t"
250	250	"pmullh %[D], %[D], %[mask] \n\t"
251	251	"pcmpgth %[Ed], %[ftmp10], %[ftmp5] \n\t"
252		− "or %[mask], %[Ed], %[Gd] \n\t"
	252	+ "por %[mask], %[Ed], %[Gd] \n\t"
253	253	"pmullh %[Ad], %[ftmp5], %[mask] \n\t"
254	254	"pmulhuh %[Ad], %[ftmp8], %[Ad] \n\t"
255	255	"pmullh %[Ad], %[Ad], %[mask] \n\t"
@@ -260,14 +260,14 @@ static void idct_column_true_mmi(uint8_t dst, int stride, int16_t input) · 4 unmodified lines
260	260	LOAD_CONST(%[ftmp8], 46341)
261	261	"psubh %[Ad], %[A], %[C] \n\t"
262	262	"pcmpgth %[Bd], %[ftmp10], %[Ad] \n\t"
263		− "or %[mask], %[Bd], %[Gd] \n\t"
	263	+ "por %[mask], %[Bd], %[Gd] \n\t"
264	264	"pmullh %[Ad], %[Ad], %[mask] \n\t"
265	265	"pmulhuh %[Ad], %[ftmp8], %[Ad] \n\t"
266	266	"pmullh %[Ad], %[Ad], %[mask] \n\t"
267	267	"paddh %[Ad], %[Ad], %[Bd] \n\t"
268	268	"psubh %[Bd], %[B], %[D] \n\t"
269	269	"pcmpgth %[Cd], %[ftmp10], %[Bd] \n\t"
270		− "or %[mask], %[Cd], %[Gd] \n\t"
	270	+ "por %[mask], %[Cd], %[Gd] \n\t"
271	271	"pmullh %[Bd], %[Bd], %[mask] \n\t"
272	272	"pmulhuh %[Bd], %[ftmp8], %[Bd] \n\t"
273	273	"pmullh %[Bd], %[Bd], %[mask] \n\t"
@@ -278,7 +278,7 @@ static void idct_column_true_mmi(uint8_t dst, int stride, int16_t input) · 4 unmodified lines
278	278	LOAD_CONST(%[Ed], 2056)
279	279	"paddh %[A], %[ftmp0], %[ftmp4] \n\t"
280	280	"pcmpgth %[B], %[ftmp10], %[A] \n\t"
281		− "or %[mask], %[B], %[Gd] \n\t"
	281	+ "por %[mask], %[B], %[Gd] \n\t"
282	282	"pmullh %[A], %[A], %[mask] \n\t"
283	283	"pmulhuh %[A], %[ftmp8], %[A] \n\t"
284	284	"pmullh %[A], %[A], %[mask] \n\t"
@@ -286,7 +286,7 @@ static void idct_column_true_mmi(uint8_t dst, int stride, int16_t input) · 1 unmodified line
286	286	"paddh %[A], %[A], %[Ed] \n\t"
287	287	"psubh %[B], %[ftmp0], %[ftmp4] \n\t"
288	288	"pcmpgth %[C], %[ftmp10], %[B] \n\t"
289		− "or %[mask], %[C], %[Gd] \n\t"
	289	+ "por %[mask], %[C], %[Gd] \n\t"
290	290	"pmullh %[B], %[B], %[mask] \n\t"
291	291	"pmulhuh %[B], %[ftmp8], %[B] \n\t"
292	292	"pmullh %[B], %[B], %[mask] \n\t"
@@ -297,14 +297,14 @@ static void idct_column_true_mmi(uint8_t dst, int stride, int16_t input) · 4 unmodified lines
297	297	LOAD_CONST(%[ftmp9], 25080)
298	298	"pmulhh %[C], %[ftmp9], %[ftmp6] \n\t"
299	299	"pcmpgth %[D], %[ftmp10], %[ftmp2] \n\t"
300		− "or %[mask], %[D], %[Gd] \n\t"
	300	+ "por %[mask], %[D], %[Gd] \n\t"
301	301	"pmullh %[Ed], %[ftmp2], %[mask] \n\t"
302	302	"pmulhuh %[Ed], %[ftmp8], %[Ed] \n\t"
303	303	"pmullh %[Ed], %[Ed], %[mask] \n\t"
304	304	"paddh %[C], %[C], %[Ed] \n\t"
305	305	"paddh %[C], %[C], %[D] \n\t"
306	306	"pcmpgth %[Ed], %[ftmp10], %[ftmp6] \n\t"
307		− "or %[mask], %[Ed], %[Gd] \n\t"
	307	+ "por %[mask], %[Ed], %[Gd] \n\t"
308	308	"pmullh %[D], %[ftmp6], %[mask] \n\t"
309	309	"pmulhuh %[D], %[ftmp8], %[D] \n\t"
310	310	"pmullh %[D], %[D], %[mask] \n\t"
@@ -317,12 +317,12 @@ static void idct_column_true_mmi(uint8_t dst, int stride, int16_t input) · 6 unmodified lines
317	317	"psubh %[C], %[B], %[Ad] \n\t"
318	318	"psubh %[B], %[Bd], %[D] \n\t"
319	319	"paddh %[D], %[Bd], %[D] \n\t"
320		− "or %[mask], %[ftmp1], %[ftmp2] \n\t"
321		− "or %[mask], %[mask], %[ftmp3] \n\t"
322		− "or %[mask], %[mask], %[ftmp4] \n\t"
323		− "or %[mask], %[mask], %[ftmp5] \n\t"
324		− "or %[mask], %[mask], %[ftmp6] \n\t"
325		− "or %[mask], %[mask], %[ftmp7] \n\t"
	320	+ "por %[mask], %[ftmp1], %[ftmp2] \n\t"
	321	+ "por %[mask], %[mask], %[ftmp3] \n\t"
	322	+ "por %[mask], %[mask], %[ftmp4] \n\t"
	323	+ "por %[mask], %[mask], %[ftmp5] \n\t"
	324	+ "por %[mask], %[mask], %[ftmp6] \n\t"
	325	+ "por %[mask], %[mask], %[ftmp7] \n\t"
326	326	"pcmpeqh %[mask], %[mask], %[ftmp10] \n\t"
327	327	"packushb %[mask], %[mask], %[ftmp10] \n\t"
328	328	"li %[tmp1], 0x04 \n\t"
@@ -361,7 +361,7 @@ static void idct_column_true_mmi(uint8_t dst, int stride, int16_t input) · 32 unmodified lines
361	361	"packushb %[ftmp7], %[ftmp7], %[ftmp10] \n\t"
362	362
363	363	"lwc1 %[Ed], 0x00(%[temp_value]) \n\t"
364		− "and %[Ed], %[Ed], %[mask] \n\t"
	364	+ "pand %[Ed], %[Ed], %[mask] \n\t"
365	365	"paddb %[ftmp0], %[ftmp0], %[Ed] \n\t"
366	366	"paddb %[ftmp1], %[ftmp1], %[Ed] \n\t"
367	367	"paddb %[ftmp2], %[ftmp2], %[Ed] \n\t"
@@ -412,7 +412,7 @@ static void idct_column_false_mmi(uint8_t dst, int stride, int16_t input) · 44 unmodified lines
412	412	for (int i = 0; i < 8; ++i)
413	413	temp_value[i] = (46341 * input[i << 3] + (8 << 16)) >> 20;
414	414	__asm__ volatile (
415		− "xor %[ftmp10], %[ftmp10], %[ftmp10] \n\t"
	415	+ "pxor %[ftmp10], %[ftmp10], %[ftmp10] \n\t"
416	416	"li %[tmp0], 0x02 \n\t"
417	417	"1: \n\t"
418	418	"ldc1 %[ftmp0], 0x00(%[input]) \n\t"
@@ -432,14 +432,14 @@ static void idct_column_false_mmi(uint8_t dst, int stride, int16_t input) · 13 unmodified lines
432	432	LOAD_CONST(%[Gd], 1)
433	433	"pmulhh %[A], %[ftmp9], %[ftmp7] \n\t"
434	434	"pcmpgth %[C], %[ftmp10], %[ftmp1] \n\t"
435		− "or %[mask], %[C], %[Gd] \n\t"
	435	+ "por %[mask], %[C], %[Gd] \n\t"
436	436	"pmullh %[B], %[ftmp1], %[mask] \n\t"
437	437	"pmulhuh %[B], %[ftmp8], %[B] \n\t"
438	438	"pmullh %[B], %[B], %[mask] \n\t"
439	439	"paddh %[A], %[A], %[B] \n\t"
440	440	"paddh %[A], %[A], %[C] \n\t"
441	441	"pcmpgth %[D], %[ftmp10], %[ftmp7] \n\t"
442		− "or %[mask], %[D], %[Gd] \n\t"
	442	+ "por %[mask], %[D], %[Gd] \n\t"
443	443	"pmullh %[Ad], %[ftmp7], %[mask] \n\t"
444	444	"pmulhuh %[B], %[ftmp8], %[Ad] \n\t"
445	445	"pmullh %[B], %[B], %[mask] \n\t"
@@ -450,12 +450,12 @@ static void idct_column_false_mmi(uint8_t dst, int stride, int16_t input) · 4 unmodified lines
450	450	LOAD_CONST(%[ftmp8], 54491)
451	451	LOAD_CONST(%[ftmp9], 36410)
452	452	"pcmpgth %[Ad], %[ftmp10], %[ftmp5] \n\t"
453		− "or %[mask], %[Ad], %[Gd] \n\t"
	453	+ "por %[mask], %[Ad], %[Gd] \n\t"
454	454	"pmullh %[Cd], %[ftmp5], %[mask] \n\t"
455	455	"pmulhuh %[C], %[ftmp9], %[Cd] \n\t"
456	456	"pmullh %[C], %[C], %[mask] \n\t"
457	457	"pcmpgth %[Bd], %[ftmp10], %[ftmp3] \n\t"
458		− "or %[mask], %[Bd], %[Gd] \n\t"
	458	+ "por %[mask], %[Bd], %[Gd] \n\t"
459	459	"pmullh %[D], %[ftmp3], %[mask] \n\t"
460	460	"pmulhuh %[D], %[ftmp8], %[D] \n\t"
461	461	"pmullh %[D], %[D], %[mask] \n\t"
@@ -463,12 +463,12 @@ static void idct_column_false_mmi(uint8_t dst, int stride, int16_t input) · 1 unmodified line
463	463	"paddh %[C], %[C], %[Ad] \n\t"
464	464	"paddh %[C], %[C], %[Bd] \n\t"
465	465	"pcmpgth %[Bd], %[ftmp10], %[ftmp3] \n\t"
466		− "or %[mask], %[Bd], %[Gd] \n\t"
	466	+ "por %[mask], %[Bd], %[Gd] \n\t"
467	467	"pmullh %[Cd], %[ftmp3], %[mask] \n\t"
468	468	"pmulhuh %[D], %[ftmp9], %[Cd] \n\t"
469	469	"pmullh %[D], %[D], %[mask] \n\t"
470	470	"pcmpgth %[Ed], %[ftmp10], %[ftmp5] \n\t"
471		− "or %[mask], %[Ed], %[Gd] \n\t"
	471	+ "por %[mask], %[Ed], %[Gd] \n\t"
472	472	"pmullh %[Ad], %[ftmp5], %[mask] \n\t"
473	473	"pmulhuh %[Ad], %[ftmp8], %[Ad] \n\t"
474	474	"pmullh %[Ad], %[Ad], %[mask] \n\t"
@@ -479,14 +479,14 @@ static void idct_column_false_mmi(uint8_t dst, int stride, int16_t input) · 4 unmodified lines
479	479	LOAD_CONST(%[ftmp8], 46341)
480	480	"psubh %[Ad], %[A], %[C] \n\t"
481	481	"pcmpgth %[Bd], %[ftmp10], %[Ad] \n\t"
482		− "or %[mask], %[Bd], %[Gd] \n\t"
	482	+ "por %[mask], %[Bd], %[Gd] \n\t"
483	483	"pmullh %[Ad], %[Ad], %[mask] \n\t"
484	484	"pmulhuh %[Ad], %[ftmp8], %[Ad] \n\t"
485	485	"pmullh %[Ad], %[Ad], %[mask] \n\t"
486	486	"paddh %[Ad], %[Ad], %[Bd] \n\t"
487	487	"psubh %[Bd], %[B], %[D] \n\t"
488	488	"pcmpgth %[Cd], %[ftmp10], %[Bd] \n\t"
489		− "or %[mask], %[Cd], %[Gd] \n\t"
	489	+ "por %[mask], %[Cd], %[Gd] \n\t"
490	490	"pmullh %[Bd], %[Bd], %[mask] \n\t"
491	491	"pmulhuh %[Bd], %[ftmp8], %[Bd] \n\t"
492	492	"pmullh %[Bd], %[Bd], %[mask] \n\t"
@@ -497,7 +497,7 @@ static void idct_column_false_mmi(uint8_t dst, int stride, int16_t input) · 4 unmodified lines
497	497	LOAD_CONST(%[Ed], 8)
498	498	"paddh %[A], %[ftmp0], %[ftmp4] \n\t"
499	499	"pcmpgth %[B], %[ftmp10], %[A] \n\t"
500		− "or %[mask], %[B], %[Gd] \n\t"
	500	+ "por %[mask], %[B], %[Gd] \n\t"
501	501	"pmullh %[A], %[A], %[mask] \n\t"
502	502	"pmulhuh %[A], %[ftmp8], %[A] \n\t"
503	503	"pmullh %[A], %[A], %[mask] \n\t"
@@ -505,7 +505,7 @@ static void idct_column_false_mmi(uint8_t dst, int stride, int16_t input) · 1 unmodified line
505	505	"paddh %[A], %[A], %[Ed] \n\t"
506	506	"psubh %[B], %[ftmp0], %[ftmp4] \n\t"
507	507	"pcmpgth %[C], %[ftmp10], %[B] \n\t"
508		− "or %[mask], %[C], %[Gd] \n\t"
	508	+ "por %[mask], %[C], %[Gd] \n\t"
509	509	"pmullh %[B], %[B], %[mask] \n\t"
510	510	"pmulhuh %[B], %[ftmp8], %[B] \n\t"
511	511	"pmullh %[B], %[B], %[mask] \n\t"
@@ -516,14 +516,14 @@ static void idct_column_false_mmi(uint8_t dst, int stride, int16_t input) · 4 unmodified lines
516	516	LOAD_CONST(%[ftmp9], 25080)
517	517	"pmulhh %[C], %[ftmp9], %[ftmp6] \n\t"
518	518	"pcmpgth %[D], %[ftmp10], %[ftmp2] \n\t"
519		− "or %[mask], %[D], %[Gd] \n\t"
	519	+ "por %[mask], %[D], %[Gd] \n\t"
520	520	"pmullh %[Ed], %[ftmp2], %[mask] \n\t"
521	521	"pmulhuh %[Ed], %[ftmp8], %[Ed] \n\t"
522	522	"pmullh %[Ed], %[Ed], %[mask] \n\t"
523	523	"paddh %[C], %[C], %[Ed] \n\t"
524	524	"paddh %[C], %[C], %[D] \n\t"
525	525	"pcmpgth %[Ed], %[ftmp10], %[ftmp6] \n\t"
526		− "or %[mask], %[Ed], %[Gd] \n\t"
	526	+ "por %[mask], %[Ed], %[Gd] \n\t"
527	527	"pmullh %[D], %[ftmp6], %[mask] \n\t"
528	528	"pmulhuh %[D], %[ftmp8], %[D] \n\t"
529	529	"pmullh %[D], %[D], %[mask] \n\t"
@@ -536,12 +536,12 @@ static void idct_column_false_mmi(uint8_t dst, int stride, int16_t input) · 6 unmodified lines
536	536	"psubh %[C], %[B], %[Ad] \n\t"
537	537	"psubh %[B], %[Bd], %[D] \n\t"
538	538	"paddh %[D], %[Bd], %[D] \n\t"
539		− "or %[mask], %[ftmp1], %[ftmp2] \n\t"
540		− "or %[mask], %[mask], %[ftmp3] \n\t"
541		− "or %[mask], %[mask], %[ftmp4] \n\t"
542		− "or %[mask], %[mask], %[ftmp5] \n\t"
543		− "or %[mask], %[mask], %[ftmp6] \n\t"
544		− "or %[mask], %[mask], %[ftmp7] \n\t"
	539	+ "por %[mask], %[ftmp1], %[ftmp2] \n\t"
	540	+ "por %[mask], %[mask], %[ftmp3] \n\t"
	541	+ "por %[mask], %[mask], %[ftmp4] \n\t"
	542	+ "por %[mask], %[mask], %[ftmp5] \n\t"
	543	+ "por %[mask], %[mask], %[ftmp6] \n\t"
	544	+ "por %[mask], %[mask], %[ftmp7] \n\t"
545	545	"pcmpeqh %[mask], %[mask], %[ftmp10] \n\t"
546	546	"li %[tmp1], 0x04 \n\t"
547	547	"dmtc1 %[tmp1], %[ftmp8] \n\t"
@@ -587,16 +587,16 @@ static void idct_column_false_mmi(uint8_t dst, int stride, int16_t input) · 39 unmodified lines
587	587	"punpcklbh %[Cd], %[Cd], %[ftmp10] \n\t"
588	588	"punpcklbh %[Dd], %[Dd], %[ftmp10] \n\t"
589	589	"ldc1 %[Ed], 0x00(%[temp_value]) \n\t"
590		− "and %[Ed], %[Ed], %[mask] \n\t"
591		− "nor %[mask], %[mask], %[mask] \n\t"
592		− "and %[ftmp0], %[ftmp0], %[mask] \n\t"
593		− "and %[ftmp1], %[ftmp1], %[mask] \n\t"
594		− "and %[ftmp2], %[ftmp2], %[mask] \n\t"
595		− "and %[ftmp3], %[ftmp3], %[mask] \n\t"
596		− "and %[ftmp4], %[ftmp4], %[mask] \n\t"
597		− "and %[ftmp5], %[ftmp5], %[mask] \n\t"
598		− "and %[ftmp6], %[ftmp6], %[mask] \n\t"
599		− "and %[ftmp7], %[ftmp7], %[mask] \n\t"
	590	+ "pand %[Ed], %[Ed], %[mask] \n\t"
	591	+ "pnor %[mask], %[mask], %[mask] \n\t"
	592	+ "pand %[ftmp0], %[ftmp0], %[mask] \n\t"
	593	+ "pand %[ftmp1], %[ftmp1], %[mask] \n\t"
	594	+ "pand %[ftmp2], %[ftmp2], %[mask] \n\t"
	595	+ "pand %[ftmp3], %[ftmp3], %[mask] \n\t"
	596	+ "pand %[ftmp4], %[ftmp4], %[mask] \n\t"
	597	+ "pand %[ftmp5], %[ftmp5], %[mask] \n\t"
	598	+ "pand %[ftmp6], %[ftmp6], %[mask] \n\t"
	599	+ "pand %[ftmp7], %[ftmp7], %[mask] \n\t"
600	600	"paddh %[ftmp0], %[ftmp0], %[A] \n\t"
601	601	"paddh %[ftmp1], %[ftmp1], %[B] \n\t"
602	602	"paddh %[ftmp2], %[ftmp2], %[C] \n\t"
@@ -689,7 +689,7 @@ void ff_vp3_idct_dc_add_mmi(uint8_t dest, ptrdiff_t line_size, int16_t block) · 86 unmodified lines
689	689	double ftmp[7];
690	690	uint64_t tmp;
691	691	__asm__ volatile (
692		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	692	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
693	693	"mtc1 %[dc], %[ftmp5] \n\t"
694	694	"pshufh %[ftmp5], %[ftmp5], %[ftmp0] \n\t"
695	695	"li %[tmp0], 0x08 \n\t"
@@ -734,10 +734,10 @@ void ff_put_no_rnd_pixels_l2_mmi(uint8_t dst, const uint8_t src1, · 38 unmodified lines
734	734	"gsldrc1 %[ftmp1], 0x00(%[src1]) \n\t"
735	735	"gsldlc1 %[ftmp2], 0x07(%[src2]) \n\t"
736	736	"gsldrc1 %[ftmp2], 0x00(%[src2]) \n\t"
737		− "xor %[ftmp3], %[ftmp1], %[ftmp2] \n\t"
738		− "and %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
	737	+ "pxor %[ftmp3], %[ftmp1], %[ftmp2] \n\t"
	738	+ "pand %[ftmp3], %[ftmp3], %[ftmp4] \n\t"
739	739	"psrlw %[ftmp3], %[ftmp3], %[ftmp5] \n\t"
740		− "and %[ftmp6], %[ftmp1], %[ftmp2] \n\t"
	740	+ "pand %[ftmp6], %[ftmp1], %[ftmp2] \n\t"
741	741	"paddw %[ftmp3], %[ftmp3], %[ftmp6] \n\t"
742	742	"sdc1 %[ftmp3], 0x00(%[dst]) \n\t"
743	743	PTR_ADDU "%[src1], %[src1], %[stride] \n\t"

libavcodec/mips/vp8dsp_mmi.c+40−40

@@ -38,10 +38,10 @@ · 37 unmodified lines
38	38	"pcmpeqb %[db_1], "#src1", "#src2" \n\t" \
39	39	"pmaxub %[db_2], "#src1", "#src2" \n\t" \
40	40	"pcmpeqb %[db_2], %[db_2], "#src1" \n\t" \
41		− "xor "#dst", %[db_2], %[db_1] \n\t"
	41	+ "pxor "#dst", %[db_2], %[db_1] \n\t"
42	42
43	43	#define MMI_BTOH(dst_l, dst_r, src) \
44		− "xor %[db_1], %[db_1], %[db_1] \n\t" \
	44	+ "pxor %[db_1], %[db_1], %[db_1] \n\t" \
45	45	"pcmpgtb %[db_2], %[db_1], "#src" \n\t" \
46	46	"punpcklbh "#dst_r", "#src", %[db_2] \n\t" \
47	47	"punpckhbh "#dst_l", "#src", %[db_2] \n\t"
@@ -84,17 +84,17 @@ · 36 unmodified lines
84	84	"punpcklwd %[ftmp3], %[ftmp3], %[ftmp3] \n\t" \
85	85	MMI_PCMPGTUB(%[mask], %[mask], %[ftmp3]) \
86	86	"pcmpeqw %[ftmp3], %[ftmp3], %[ftmp3] \n\t" \
87		− "xor %[mask], %[mask], %[ftmp3] \n\t" \
	87	+ "pxor %[mask], %[mask], %[ftmp3] \n\t" \
88	88	/* VP8_MBFILTER */ \
89	89	"li %[tmp0], 0x80808080 \n\t" \
90	90	"dmtc1 %[tmp0], %[ftmp7] \n\t" \
91	91	"punpcklwd %[ftmp7], %[ftmp7], %[ftmp7] \n\t" \
92		− "xor %[p2], %[p2], %[ftmp7] \n\t" \
93		− "xor %[p1], %[p1], %[ftmp7] \n\t" \
94		− "xor %[p0], %[p0], %[ftmp7] \n\t" \
95		− "xor %[q0], %[q0], %[ftmp7] \n\t" \
96		− "xor %[q1], %[q1], %[ftmp7] \n\t" \
97		− "xor %[q2], %[q2], %[ftmp7] \n\t" \
	92	+ "pxor %[p2], %[p2], %[ftmp7] \n\t" \
	93	+ "pxor %[p1], %[p1], %[ftmp7] \n\t" \
	94	+ "pxor %[p0], %[p0], %[ftmp7] \n\t" \
	95	+ "pxor %[q0], %[q0], %[ftmp7] \n\t" \
	96	+ "pxor %[q1], %[q1], %[ftmp7] \n\t" \
	97	+ "pxor %[q2], %[q2], %[ftmp7] \n\t" \
98	98	"psubsb %[ftmp4], %[p1], %[q1] \n\t" \
99	99	"psubb %[ftmp5], %[q0], %[p0] \n\t" \
100	100	MMI_BTOH(%[ftmp1], %[ftmp0], %[ftmp5]) \
@@ -109,8 +109,8 @@ · 8 unmodified lines
109	109	"paddh %[ftmp1], %[ftmp3], %[ftmp1] \n\t" \
110	110	/* Combine left and right part */ \
111	111	"packsshb %[ftmp1], %[ftmp0], %[ftmp1] \n\t" \
112		− "and %[ftmp1], %[ftmp1], %[mask] \n\t" \
113		− "and %[ftmp2], %[ftmp1], %[hev] \n\t" \
	112	+ "pand %[ftmp1], %[ftmp1], %[mask] \n\t" \
	113	+ "pand %[ftmp2], %[ftmp1], %[hev] \n\t" \
114	114	"li %[tmp0], 0x04040404 \n\t" \
115	115	"dmtc1 %[tmp0], %[ftmp0] \n\t" \
116	116	"punpcklwd %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
@@ -129,8 +129,8 @@ · 12 unmodified lines
129	129	"paddsb %[p0], %[p0], %[ftmp4] \n\t" \
130	130	/* filt_val &= ~hev */ \
131	131	"pcmpeqw %[ftmp0], %[ftmp0], %[ftmp0] \n\t" \
132		− "xor %[hev], %[hev], %[ftmp0] \n\t" \
133		− "and %[ftmp1], %[ftmp1], %[hev] \n\t" \
	132	+ "pxor %[hev], %[hev], %[ftmp0] \n\t" \
	133	+ "pand %[ftmp1], %[ftmp1], %[hev] \n\t" \
134	134	MMI_BTOH(%[ftmp5], %[ftmp6], %[ftmp1]) \
135	135	"li %[tmp0], 0x07 \n\t" \
136	136	"dmtc1 %[tmp0], %[ftmp2] \n\t" \
@@ -151,9 +151,9 @@ · 14 unmodified lines
151	151	/* Combine left and right part */ \
152	152	"packsshb %[ftmp4], %[ftmp3], %[ftmp4] \n\t" \
153	153	"psubsb %[q0], %[q0], %[ftmp4] \n\t" \
154		− "xor %[q0], %[q0], %[ftmp7] \n\t" \
	154	+ "pxor %[q0], %[q0], %[ftmp7] \n\t" \
155	155	"paddsb %[p0], %[p0], %[ftmp4] \n\t" \
156		− "xor %[p0], %[p0], %[ftmp7] \n\t" \
	156	+ "pxor %[p0], %[p0], %[ftmp7] \n\t" \
157	157	"li %[tmp0], 0x00120012 \n\t" \
158	158	"dmtc1 %[tmp0], %[ftmp1] \n\t" \
159	159	"punpcklwd %[ftmp1], %[ftmp1], %[ftmp1] \n\t" \
@@ -168,9 +168,9 @@ · 8 unmodified lines
168	168	/* Combine left and right part */ \
169	169	"packsshb %[ftmp4], %[ftmp3], %[ftmp4] \n\t" \
170	170	"psubsb %[q1], %[q1], %[ftmp4] \n\t" \
171		− "xor %[q1], %[q1], %[ftmp7] \n\t" \
	171	+ "pxor %[q1], %[q1], %[ftmp7] \n\t" \
172	172	"paddsb %[p1], %[p1], %[ftmp4] \n\t" \
173		− "xor %[p1], %[p1], %[ftmp7] \n\t" \
	173	+ "pxor %[p1], %[p1], %[ftmp7] \n\t" \
174	174	"li %[tmp0], 0x03 \n\t" \
175	175	"dmtc1 %[tmp0], %[ftmp1] \n\t" \
176	176	/* Right part */ \
@@ -186,9 +186,9 @@ · 9 unmodified lines
186	186	/* Combine left and right part */ \
187	187	"packsshb %[ftmp4], %[ftmp3], %[ftmp4] \n\t" \
188	188	"psubsb %[q2], %[q2], %[ftmp4] \n\t" \
189		− "xor %[q2], %[q2], %[ftmp7] \n\t" \
	189	+ "pxor %[q2], %[q2], %[ftmp7] \n\t" \
190	190	"paddsb %[p2], %[p2], %[ftmp4] \n\t" \
191		− "xor %[p2], %[p2], %[ftmp7] \n\t"
	191	+ "pxor %[p2], %[p2], %[ftmp7] \n\t"
192	192
193	193	#define PUT_VP8_EPEL4_H6_MMI(src, dst) \
194	194	MMI_ULWC1(%[ftmp1], src, 0x00) \
@@ -1021,7 +1021,7 @@ void ff_vp8_luma_dc_wht_mmi(int16_t block[4][4][16], int16_t dc[16]) · 826 unmodified lines
1021	1021	block[3][3][0] = (dc[12] - dc[15] + 3 - dc[13] + dc[14]) >> 3;
1022	1022
1023	1023	__asm__ volatile (
1024		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1024	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1025	1025	MMI_SDC1(%[ftmp0], %[dc], 0x00)
1026	1026	MMI_SDC1(%[ftmp0], %[dc], 0x08)
1027	1027	MMI_SDC1(%[ftmp0], %[dc], 0x10)
@@ -1136,7 +1136,7 @@ void ff_vp8_idct_add_mmi(uint8_t *dst, int16_t block[16], ptrdiff_t stride) · 108 unmodified lines
1136	1136	DECLARE_VAR_ALL64;
1137	1137
1138	1138	__asm__ volatile (
1139		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1139	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1140	1140	MMI_LDC1(%[ftmp1], %[block], 0x00)
1141	1141	MMI_LDC1(%[ftmp2], %[block], 0x08)
1142	1142	MMI_LDC1(%[ftmp3], %[block], 0x10)
@@ -1302,7 +1302,7 @@ void ff_vp8_idct_dc_add_mmi(uint8_t *dst, int16_t block[16], ptrdiff_t stride) · 159 unmodified lines
1302	1302	block[0] = 0;
1303	1303
1304	1304	__asm__ volatile (
1305		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1305	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1306	1306	"mtc1 %[dc], %[ftmp5] \n\t"
1307	1307	MMI_LWC1(%[ftmp1], %[dst0], 0x00)
1308	1308	MMI_LWC1(%[ftmp2], %[dst1], 0x00)
@@ -1618,7 +1618,7 @@ void ff_put_vp8_epel16_h4_mmi(uint8_t dst, ptrdiff_t dststride, uint8_t src, · 309 unmodified lines
1618	1618	dst[15] = cm[(filter[2] * src[15] - filter[1] * src[14] + filter[3] * src[16] - filter[4] * src[17] + 64) >> 7];
1619	1619	*/
1620	1620	__asm__ volatile (
1621		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1621	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1622	1622	"li %[tmp0], 0x07 \n\t"
1623	1623	"mtc1 %[tmp0], %[ftmp4] \n\t"
1624	1624
@@ -1685,7 +1685,7 @@ void ff_put_vp8_epel8_h4_mmi(uint8_t dst, ptrdiff_t dststride, uint8_t src, · 60 unmodified lines
1685	1685	dst[7] = cm[(filter[2] * src[7] - filter[1] * src[ 6] + filter[3] * src[8] - filter[4] * src[9] + 64) >> 7];
1686	1686	*/
1687	1687	__asm__ volatile (
1688		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1688	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1689	1689	"li %[tmp0], 0x07 \n\t"
1690	1690	"mtc1 %[tmp0], %[ftmp4] \n\t"
1691	1691
@@ -1742,7 +1742,7 @@ void ff_put_vp8_epel4_h4_mmi(uint8_t dst, ptrdiff_t dststride, uint8_t src, · 50 unmodified lines
1742	1742	dst[3] = cm[(filter[2] * src[3] - filter[1] * src[ 2] + filter[3] * src[4] - filter[4] * src[5] + 64) >> 7];
1743	1743	*/
1744	1744	__asm__ volatile (
1745		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1745	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1746	1746	"li %[tmp0], 0x07 \n\t"
1747	1747	"mtc1 %[tmp0], %[ftmp4] \n\t"
1748	1748
@@ -1811,7 +1811,7 @@ void ff_put_vp8_epel16_h6_mmi(uint8_t dst, ptrdiff_t dststride, uint8_t src, · 62 unmodified lines
1811	1811	dst[15] = cm[(filter[2]src[15] - filter[1]src[14] + filter[0]src[13] + filter[3]src[16] - filter[4]src[17] + filter[5]src[18] + 64) >> 7];
1812	1812	*/
1813	1813	__asm__ volatile (
1814		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1814	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1815	1815	"li %[tmp0], 0x07 \n\t"
1816	1816	"mtc1 %[tmp0], %[ftmp4] \n\t"
1817	1817
@@ -1879,7 +1879,7 @@ void ff_put_vp8_epel8_h6_mmi(uint8_t dst, ptrdiff_t dststride, uint8_t src, · 61 unmodified lines
1879	1879	dst[7] = cm[(filter[2]src[7] - filter[1]src[ 6] + filter[0]src[ 5] + filter[3]src[8] - filter[4]src[9] + filter[5]src[10] + 64) >> 7];
1880	1880	*/
1881	1881	__asm__ volatile (
1882		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1882	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1883	1883	"li %[tmp0], 0x07 \n\t"
1884	1884	"mtc1 %[tmp0], %[ftmp4] \n\t"
1885	1885
@@ -1937,7 +1937,7 @@ void ff_put_vp8_epel4_h6_mmi(uint8_t dst, ptrdiff_t dststride, uint8_t src, · 51 unmodified lines
1937	1937	dst[3] = cm[(filter[2]src[3] - filter[1]src[ 2] + filter[0]src[ 1] + filter[3]src[4] - filter[4]src[5] + filter[5]src[ 6] + 64) >> 7];
1938	1938	*/
1939	1939	__asm__ volatile (
1940		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	1940	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
1941	1941	"li %[tmp0], 0x07 \n\t"
1942	1942	"mtc1 %[tmp0], %[ftmp4] \n\t"
1943	1943
@@ -2007,7 +2007,7 @@ void ff_put_vp8_epel16_v4_mmi(uint8_t dst, ptrdiff_t dststride, uint8_t src, · 63 unmodified lines
2007	2007	dst[15] = cm[(filter[2] * src[15] - filter[1] * src[15-srcstride] + filter[3] * src[15+srcstride] - filter[4] * src[15+2*srcstride] + 64) >> 7];
2008	2008	*/
2009	2009	__asm__ volatile (
2010		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	2010	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
2011	2011	"li %[tmp0], 0x07 \n\t"
2012	2012	"mtc1 %[tmp0], %[ftmp4] \n\t"
2013	2013
@@ -2076,7 +2076,7 @@ void ff_put_vp8_epel8_v4_mmi(uint8_t dst, ptrdiff_t dststride, uint8_t src, · 62 unmodified lines
2076	2076	dst[7] = cm[(filter[2] * src[7] - filter[1] * src[7-srcstride] + filter[3] * src[7+srcstride] - filter[4] * src[7+2*srcstride] + 64) >> 7];
2077	2077	*/
2078	2078	__asm__ volatile (
2079		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	2079	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
2080	2080	"li %[tmp0], 0x07 \n\t"
2081	2081	"mtc1 %[tmp0], %[ftmp4] \n\t"
2082	2082
@@ -2135,7 +2135,7 @@ void ff_put_vp8_epel4_v4_mmi(uint8_t dst, ptrdiff_t dststride, uint8_t src, · 52 unmodified lines
2135	2135	dst[3] = cm[(filter[2] * src[3] - filter[1] * src[3-srcstride] + filter[3] * src[3+srcstride] - filter[4] * src[3+2*srcstride] + 64) >> 7];
2136	2136	*/
2137	2137	__asm__ volatile (
2138		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	2138	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
2139	2139	"li %[tmp0], 0x07 \n\t"
2140	2140	"mtc1 %[tmp0], %[ftmp4] \n\t"
2141	2141
@@ -2205,7 +2205,7 @@ void ff_put_vp8_epel16_v6_mmi(uint8_t dst, ptrdiff_t dststride, uint8_t src, · 63 unmodified lines
2205	2205	dst[15] = cm[(filter[2]src[15] - filter[1]src[15-srcstride] + filter[0]src[15-2srcstride] + filter[3]src[15+srcstride] - filter[4]src[15+2srcstride] + filter[5]src[15+3*srcstride] + 64) >> 7];
2206	2206	*/
2207	2207	__asm__ volatile (
2208		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	2208	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
2209	2209	"li %[tmp0], 0x07 \n\t"
2210	2210	"mtc1 %[tmp0], %[ftmp4] \n\t"
2211	2211
@@ -2275,7 +2275,7 @@ void ff_put_vp8_epel8_v6_mmi(uint8_t dst, ptrdiff_t dststride, uint8_t src, · 63 unmodified lines
2275	2275	dst[7] = cm[(filter[2]src[7] - filter[1]src[7-srcstride] + filter[0]src[7-2srcstride] + filter[3]src[7+srcstride] - filter[4]src[7+2srcstride] + filter[5]src[7+3*srcstride] + 64) >> 7];
2276	2276	*/
2277	2277	__asm__ volatile (
2278		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	2278	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
2279	2279	"li %[tmp0], 0x07 \n\t"
2280	2280	"mtc1 %[tmp0], %[ftmp4] \n\t"
2281	2281
@@ -2335,7 +2335,7 @@ void ff_put_vp8_epel4_v6_mmi(uint8_t dst, ptrdiff_t dststride, uint8_t src, · 53 unmodified lines
2335	2335	dst[3] = cm[(filter[2]src[3] - filter[1]src[3-srcstride] + filter[0]src[3-2srcstride] + filter[3]src[3+srcstride] - filter[4]src[3+2srcstride] + filter[5]src[3+3*srcstride] + 64) >> 7];
2336	2336	*/
2337	2337	__asm__ volatile (
2338		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	2338	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
2339	2339	"li %[tmp0], 0x07 \n\t"
2340	2340	"mtc1 %[tmp0], %[ftmp4] \n\t"
2341	2341
@@ -2873,7 +2873,7 @@ void ff_put_vp8_bilinear16_h_mmi(uint8_t dst, ptrdiff_t dstride, uint8_t src, · 531 unmodified lines
2873	2873	dst[15] = (a * src[15] + b * src[16] + 4) >> 3;
2874	2874	*/
2875	2875	__asm__ volatile (
2876		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	2876	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
2877	2877	"li %[tmp0], 0x03 \n\t"
2878	2878	"mtc1 %[tmp0], %[ftmp4] \n\t"
2879	2879	"pshufh %[a], %[a], %[ftmp0] \n\t"
@@ -2940,7 +2940,7 @@ void ff_put_vp8_bilinear16_v_mmi(uint8_t dst, ptrdiff_t dstride, uint8_t src, · 60 unmodified lines
2940	2940	dst[7] = (c * src[7] + d * src[7 + sstride] + 4) >> 3;
2941	2941	*/
2942	2942	__asm__ volatile (
2943		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	2943	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
2944	2944	"li %[tmp0], 0x03 \n\t"
2945	2945	"mtc1 %[tmp0], %[ftmp4] \n\t"
2946	2946	"pshufh %[c], %[c], %[ftmp0] \n\t"
@@ -3041,7 +3041,7 @@ void ff_put_vp8_bilinear8_h_mmi(uint8_t dst, ptrdiff_t dstride, uint8_t src, · 94 unmodified lines
3041	3041	dst[7] = (a * src[7] + b * src[8] + 4) >> 3;
3042	3042	*/
3043	3043	__asm__ volatile (
3044		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	3044	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
3045	3045	"li %[tmp0], 0x03 \n\t"
3046	3046	"mtc1 %[tmp0], %[ftmp4] \n\t"
3047	3047	"pshufh %[a], %[a], %[ftmp0] \n\t"
@@ -3102,7 +3102,7 @@ void ff_put_vp8_bilinear8_v_mmi(uint8_t dst, ptrdiff_t dstride, uint8_t src, · 54 unmodified lines
3102	3102	dst[7] = (c * src[7] + d * src[7 + sstride] + 4) >> 3;
3103	3103	*/
3104	3104	__asm__ volatile (
3105		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	3105	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
3106	3106	"li %[tmp0], 0x03 \n\t"
3107	3107	"mtc1 %[tmp0], %[ftmp4] \n\t"
3108	3108	"pshufh %[c], %[c], %[ftmp0] \n\t"
@@ -3194,7 +3194,7 @@ void ff_put_vp8_bilinear4_h_mmi(uint8_t dst, ptrdiff_t dstride, uint8_t src, · 85 unmodified lines
3194	3194	dst[3] = (a * src[3] + b * src[4] + 4) >> 3;
3195	3195	*/
3196	3196	__asm__ volatile (
3197		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	3197	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
3198	3198	"li %[tmp0], 0x03 \n\t"
3199	3199	"mtc1 %[tmp0], %[ftmp4] \n\t"
3200	3200	"pshufh %[a], %[a], %[ftmp0] \n\t"
@@ -3252,7 +3252,7 @@ void ff_put_vp8_bilinear4_v_mmi(uint8_t dst, ptrdiff_t dstride, uint8_t src, · 51 unmodified lines
3252	3252	dst[3] = (c * src[3] + d * src[3 + sstride] + 4) >> 3;
3253	3253	*/
3254	3254	__asm__ volatile (
3255		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	3255	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
3256	3256	"li %[tmp0], 0x03 \n\t"
3257	3257	"mtc1 %[tmp0], %[ftmp4] \n\t"
3258	3258	"pshufh %[c], %[c], %[ftmp0] \n\t"

libavcodec/mips/vp9_mc_mmi.c+5−5

@@ -82,7 +82,7 @@ static void convolve_horiz_mmi(const uint8_t *src, int32_t src_stride, · 81 unmodified lines
82	82	dst_stride -= w;
83	83	__asm__ volatile (
84	84	"move %[tmp1], %[width] \n\t"
85		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	85	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
86	86	"gsldlc1 %[filter1], 0x03(%[filter]) \n\t"
87	87	"gsldrc1 %[filter1], 0x00(%[filter]) \n\t"
88	88	"gsldlc1 %[filter2], 0x0b(%[filter]) \n\t"
@@ -157,7 +157,7 @@ static void convolve_vert_mmi(const uint8_t *src, int32_t src_stride, · 68 unmodified lines
157	157	dst_stride -= w;
158	158
159	159	__asm__ volatile (
160		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	160	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
161	161	"gsldlc1 %[ftmp4], 0x03(%[filter]) \n\t"
162	162	"gsldrc1 %[ftmp4], 0x00(%[filter]) \n\t"
163	163	"gsldlc1 %[ftmp5], 0x0b(%[filter]) \n\t"
@@ -253,7 +253,7 @@ static void convolve_avg_horiz_mmi(const uint8_t *src, int32_t src_stride, · 89 unmodified lines
253	253
254	254	__asm__ volatile (
255	255	"move %[tmp1], %[width] \n\t"
256		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	256	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
257	257	"gsldlc1 %[filter1], 0x03(%[filter]) \n\t"
258	258	"gsldrc1 %[filter1], 0x00(%[filter]) \n\t"
259	259	"gsldlc1 %[filter2], 0x0b(%[filter]) \n\t"
@@ -339,7 +339,7 @@ static void convolve_avg_vert_mmi(const uint8_t *src, int32_t src_stride, · 79 unmodified lines
339	339	dst_stride -= w;
340	340
341	341	__asm__ volatile (
342		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	342	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
343	343	"gsldlc1 %[ftmp4], 0x03(%[filter]) \n\t"
344	344	"gsldrc1 %[ftmp4], 0x00(%[filter]) \n\t"
345	345	"gsldlc1 %[ftmp5], 0x0b(%[filter]) \n\t"
@@ -444,7 +444,7 @@ static void convolve_avg_mmi(const uint8_t *src, int32_t src_stride, · 98 unmodified lines
444	444
445	445	__asm__ volatile (
446	446	"move %[tmp1], %[width] \n\t"
447		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	447	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
448	448	"li %[tmp0], 0x10001 \n\t"
449	449	"dmtc1 %[tmp0], %[ftmp3] \n\t"
450	450	"punpcklhw %[ftmp3], %[ftmp3], %[ftmp3] \n\t"

libavcodec/mips/wmv2dsp_mmi.c+1−1

@@ -106,7 +106,7 @@ void ff_wmv2_idct_add_mmi(uint8_t dest, ptrdiff_t line_size, int16_t block) · 105 unmodified lines
106	106	wmv2_idct_col_mmi(block + i);
107	107
108	108	__asm__ volatile (
109		− "xor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
	109	+ "pxor %[ftmp0], %[ftmp0], %[ftmp0] \n\t"
110	110
111	111	// low 4 loop
112	112	MMI_LDC1(%[ftmp1], %[block], 0x00)

libavcodec/mips: Fix specification of instruction name

Keyboard shortcuts

General

Issues and pull requests

Issues, pull requests and Inbox

Repository code