History

x86: Update x86inc.asm

Make things up-to-date with upstream. https://code.videolan.org/videolan/x86inc.asm

Henrik Gramner committed 2 years ago

Commit afa471d1 parent 782c4df12 files · +519 −270Browse files

Showing the changes to libavcodec (10 of 12 files). Show all

Unified · Split

10 files
  1. libavcodec/x86/celt_pvq_search.asm +2 −2
  2. libavcodec/x86/h264_chromamc.asm +5 −5
  3. libavcodec/x86/h264_idct.asm +3 −3
  4. libavcodec/x86/h264_intrapred.asm +6 −9
  5. libavcodec/x86/hevc_mc.asm +8 −8
  6. libavcodec/x86/rv40dsp.asm +6 −6
  7. libavcodec/x86/sbrdsp.asm +3 −3
  8. libavcodec/x86/vp8dsp.asm +15 −15
  9. libavcodec/x86/vp9itxfm.asm +2 −0
  10. libavcodec/x86/vp9itxfm_16bpp.asm +7 −5
libavcodec/x86/celt_pvq_search.asm+2−2
@@ -74,7 +74,7 @@ SECTION .text · 73 unmodified lines
74; "movaps m0, [r5 + r4]" if PIC is enabled74; "movaps m0, [r5 + r4]" if PIC is enabled
75; "movaps m0, [constant_name + r4]" if texrel are used75; "movaps m0, [constant_name + r4]" if texrel are used
76%macro SET_PIC_BASE 3; reg, const_label76%macro SET_PIC_BASE 3; reg, const_label
77%ifdef PIC77%if PIC
78 %{1} %2, [%3] ; lea r5, [rip+const]78 %{1} %2, [%3] ; lea r5, [rip+const]
79 %define pic_base_%3 %279 %define pic_base_%3 %2
80%else80%else
@@ -195,7 +195,7 @@ align 16 · 114 unmodified lines
195; PIC relative addressing. Use this195; PIC relative addressing. Use this
196; to count it in cglobal196; to count it in cglobal
197;197;
198%ifdef PIC198%if PIC
199 %define num_pic_regs 1199 %define num_pic_regs 1
200%else200%else
201 %define num_pic_regs 0201 %define num_pic_regs 0
libavcodec/x86/h264_chromamc.asm+5−5
@@ -91,7 +91,7 @@ SECTION .text · 90 unmodified lines
91 91
92%macro chroma_mc8_mmx_func 2-392%macro chroma_mc8_mmx_func 2-3
93%ifidn %2, rv4093%ifidn %2, rv40
94%ifdef PIC94%if PIC
95%define rnd_1d_rv40 r895%define rnd_1d_rv40 r8
96%define rnd_2d_rv40 r896%define rnd_2d_rv40 r8
97%define extra_regs 297%define extra_regs 2
@@ -147,7 +147,7 @@ cglobal %1_%2_chroma_mc8%3, 6, 7 + extra_regs, 0 · 49 unmodified lines
147 or r4d, r5d ; x + y147 or r4d, r5d ; x + y
148 148
149%ifidn %2, rv40149%ifidn %2, rv40
150%ifdef PIC150%if PIC
151 lea r8, [rnd_rv40_1d_tbl]151 lea r8, [rnd_rv40_1d_tbl]
152%endif152%endif
153%if ARCH_X86_64 == 0153%if ARCH_X86_64 == 0
@@ -198,7 +198,7 @@ cglobal %1_%2_chroma_mc8%3, 6, 7 + extra_regs, 0 · 44 unmodified lines
198 movd m4, r4d ; x198 movd m4, r4d ; x
199 movd m6, r5d ; y199 movd m6, r5d ; y
200%ifidn %2, rv40200%ifidn %2, rv40
201%ifdef PIC201%if PIC
202 lea r8, [rnd_rv40_2d_tbl]202 lea r8, [rnd_rv40_2d_tbl]
203%endif203%endif
204%if ARCH_X86_64 == 0204%if ARCH_X86_64 == 0
@@ -283,7 +283,7 @@ cglobal %1_%2_chroma_mc8%3, 6, 7 + extra_regs, 0 · 78 unmodified lines
283%macro chroma_mc4_mmx_func 2283%macro chroma_mc4_mmx_func 2
284%define extra_regs 0284%define extra_regs 0
285%ifidn %2, rv40285%ifidn %2, rv40
286%ifdef PIC286%if PIC
287%define extra_regs 1287%define extra_regs 1
288%endif ; PIC288%endif ; PIC
289%endif ; rv40289%endif ; rv40
@@ -301,7 +301,7 @@ cglobal %1_%2_chroma_mc4, 6, 6 + extra_regs, 0 · 11 unmodified lines
301 psubw m5, m3301 psubw m5, m3
302 302
303%ifidn %2, rv40303%ifidn %2, rv40
304%ifdef PIC304%if PIC
305 lea r6, [rnd_rv40_2d_tbl]305 lea r6, [rnd_rv40_2d_tbl]
306%define rnd_2d_rv40 r6306%define rnd_2d_rv40 r6
307%else307%else
libavcodec/x86/h264_idct.asm+3−3
@@ -42,7 +42,7 @@ scan8_mem: db 4+ 1*8, 5+ 1*8, 4+ 2*8, 5+ 2*8 · 41 unmodified lines
42 db 6+11*8, 7+11*8, 6+12*8, 7+12*842 db 6+11*8, 7+11*8, 6+12*8, 7+12*8
43 db 4+13*8, 5+13*8, 4+14*8, 5+14*843 db 4+13*8, 5+13*8, 4+14*8, 5+14*8
44 db 6+13*8, 7+13*8, 6+14*8, 7+14*844 db 6+13*8, 7+13*8, 6+14*8, 7+14*8
45%ifdef PIC45%if PIC
46%define npicregs 146%define npicregs 1
47%define scan8 picregq47%define scan8 picregq
48%else48%else
@@ -322,7 +322,7 @@ INIT_XMM sse2 · 273 unmodified lines
322cglobal h264_idct8_add4_8, 5, 8 + npicregs, 10, dst1, block_offset, block, stride, nnzc, cntr, coeff, dst2, picreg322cglobal h264_idct8_add4_8, 5, 8 + npicregs, 10, dst1, block_offset, block, stride, nnzc, cntr, coeff, dst2, picreg
323 movsxdifnidn r3, r3d323 movsxdifnidn r3, r3d
324 xor r5, r5324 xor r5, r5
325%ifdef PIC325%if PIC
326 lea picregq, [scan8_mem]326 lea picregq, [scan8_mem]
327%endif327%endif
328.nextblock:328.nextblock:
@@ -398,7 +398,7 @@ h264_idct_add8_mmx_plane: · 69 unmodified lines
398cglobal h264_idct_add8_422_8, 5, 8 + npicregs, 0, dst1, block_offset, block, stride, nnzc, cntr, coeff, dst2, picreg398cglobal h264_idct_add8_422_8, 5, 8 + npicregs, 0, dst1, block_offset, block, stride, nnzc, cntr, coeff, dst2, picreg
399; dst1, block_offset, block, stride, nnzc, cntr, coeff, dst2, picreg399; dst1, block_offset, block, stride, nnzc, cntr, coeff, dst2, picreg
400 movsxdifnidn r3, r3d400 movsxdifnidn r3, r3d
401%ifdef PIC401%if PIC
402 lea picregq, [scan8_mem]402 lea picregq, [scan8_mem]
403%endif403%endif
404%if ARCH_X86_64404%if ARCH_X86_64
libavcodec/x86/h264_intrapred.asm+6−9
@@ -1311,10 +1311,7 @@ PRED8x8L_DOWN_RIGHT · 1310 unmodified lines
1311;-----------------------------------------------------------------------------1311;-----------------------------------------------------------------------------
1312 1312
1313%macro PRED8x8L_VERTICAL_RIGHT 01313%macro PRED8x8L_VERTICAL_RIGHT 0
1314cglobal pred8x8l_vertical_right_8, 4,5,71314cglobal pred8x8l_vertical_right_8, 4,5,6
1315 ; manually spill XMM registers for Win64 because
1316 ; the code here is initialized with INIT_MMX
1317 WIN64_SPILL_XMM 7
1318 sub r0, r31315 sub r0, r3
1319 lea r4, [r0+r3*2]1316 lea r4, [r0+r3*2]
1320 movq mm0, [r0+r3*1-8]1317 movq mm0, [r0+r3*1-8]
@@ -1384,7 +1381,6 @@ cglobal pred8x8l_vertical_right_8, 4,5,7 · 63 unmodified lines
1384 movq2dq xmm4, mm61381 movq2dq xmm4, mm6
1385 pslldq xmm4, 81382 pslldq xmm4, 8
1386 por xmm0, xmm41383 por xmm0, xmm4
1387 movdqa xmm6, [pw_ff00]
1388 movdqa xmm1, xmm01384 movdqa xmm1, xmm0
1389 lea r2, [r1+r3*2]1385 lea r2, [r1+r3*2]
1390 movdqa xmm2, xmm01386 movdqa xmm2, xmm0
@@ -1394,15 +1390,16 @@ cglobal pred8x8l_vertical_right_8, 4,5,7 · 3 unmodified lines
1394 pavgb xmm2, xmm01390 pavgb xmm2, xmm0
1395INIT_XMM cpuname1391INIT_XMM cpuname
1396 PRED4x4_LOWPASS xmm4, xmm3, xmm1, xmm0, xmm51392 PRED4x4_LOWPASS xmm4, xmm3, xmm1, xmm0, xmm5
1397 pandn xmm6, xmm41393 movdqa xmm0, [pw_ff00]
1394 pandn xmm0, xmm4
1398 movdqa xmm5, xmm41395 movdqa xmm5, xmm4
1399 psrlw xmm4, 81396 psrlw xmm4, 8
1400 packuswb xmm6, xmm41397 packuswb xmm0, xmm4
1401 movhlps xmm4, xmm61398 movhlps xmm4, xmm0
1402 movhps [r0+r3*2], xmm51399 movhps [r0+r3*2], xmm5
1403 movhps [r0+r3*1], xmm21400 movhps [r0+r3*1], xmm2
1404 psrldq xmm5, 41401 psrldq xmm5, 4
1405 movss xmm5, xmm61402 movss xmm5, xmm0
1406 psrldq xmm2, 41403 psrldq xmm2, 4
1407 movss xmm2, xmm41404 movss xmm2, xmm4
1408 lea r0, [r2+r3*2]1405 lea r0, [r2+r3*2]
libavcodec/x86/hevc_mc.asm+8−8
@@ -180,7 +180,7 @@ SECTION .text · 179 unmodified lines
180%macro EPEL_FILTER 5 ; bit depth, filter index, xmma, xmmb, gprtmp180%macro EPEL_FILTER 5 ; bit depth, filter index, xmma, xmmb, gprtmp
181%if cpuflag(avx2)181%if cpuflag(avx2)
182%assign %%offset 32182%assign %%offset 32
183%ifdef PIC183%if PIC
184 lea %5q, [hevc_epel_filters_avx2_%1]184 lea %5q, [hevc_epel_filters_avx2_%1]
185 %define FILTER %5q185 %define FILTER %5q
186%else186%else
@@ -188,7 +188,7 @@ SECTION .text · 1 unmodified line
188%endif188%endif
189%else189%else
190%assign %%offset 16190%assign %%offset 16
191%ifdef PIC191%if PIC
192 lea %5q, [hevc_epel_filters_sse4_%1]192 lea %5q, [hevc_epel_filters_sse4_%1]
193 %define FILTER %5q193 %define FILTER %5q
194%else194%else
@@ -216,7 +216,7 @@ SECTION .text · 21 unmodified lines
216%define %%table hevc_epel_filters_sse4_%1216%define %%table hevc_epel_filters_sse4_%1
217%endif217%endif
218 218
219%ifdef PIC219%if PIC
220 lea r3srcq, [%%table]220 lea r3srcq, [%%table]
221 %define FILTER r3srcq221 %define FILTER r3srcq
222%else222%else
@@ -234,7 +234,7 @@ SECTION .text · 11 unmodified lines
234%else234%else
235%define %%table hevc_epel_filters_sse4_10235%define %%table hevc_epel_filters_sse4_10
236%endif236%endif
237%ifdef PIC237%if PIC
238 lea r3srcq, [%%table]238 lea r3srcq, [%%table]
239 %define FILTER r3srcq239 %define FILTER r3srcq
240%else240%else
@@ -257,7 +257,7 @@ SECTION .text · 16 unmodified lines
257%define %%table hevc_qpel_filters_sse4_%1257%define %%table hevc_qpel_filters_sse4_%1
258%endif258%endif
259 259
260%ifdef PIC260%if PIC
261 lea rfilterq, [%%table]261 lea rfilterq, [%%table]
262%else262%else
263 %define rfilterq %%table263 %define rfilterq %%table
@@ -576,7 +576,7 @@ SECTION .text · 312 unmodified lines
576%define %%table hevc_qpel_filters_sse4_%2576%define %%table hevc_qpel_filters_sse4_%2
577%endif577%endif
578 578
579%ifdef PIC579%if PIC
580 lea rfilterq, [%%table]580 lea rfilterq, [%%table]
581%else581%else
582 %define rfilterq %%table582 %define rfilterq %%table
@@ -1288,7 +1288,7 @@ HEVC_PUT_HEVC_QPEL_HV 16, 10 · 705 unmodified lines
1288%assign %%offset 41288%assign %%offset 4
1289 dec %2q1289 dec %2q
1290 shl %2q, 31290 shl %2q, 3
1291%ifdef PIC1291%if PIC
1292 lea %5q, [%%table]1292 lea %5q, [%%table]
1293 %define FILTER %5q1293 %define FILTER %5q
1294%else1294%else
@@ -1365,7 +1365,7 @@ cglobal hevc_put_hevc_qpel_hv%1_%2, 6, 7, 27, dst, src, srcstride, height, mx, m · 70 unmodified lines
1365 sub myq, 11365 sub myq, 1
1366 shl myq, 51366 shl myq, 5
1367%define %%table hevc_qpel_filters_avx512icl_v_%11367%define %%table hevc_qpel_filters_avx512icl_v_%1
1368%ifdef PIC1368%if PIC
1369 lea tmpq, [%%table]1369 lea tmpq, [%%table]
1370 %define FILTER tmpq1370 %define FILTER tmpq
1371%else1371%else
libavcodec/x86/rv40dsp.asm+6−6
@@ -51,7 +51,7 @@ sixtap_filter_v_m: times 8 dw 1 · 50 unmodified lines
51 times 8 dw 2051 times 8 dw 20
52 times 8 dw 5252 times 8 dw 52
53 53
54%ifdef PIC54%if PIC
55%define sixtap_filter_hw picregq55%define sixtap_filter_hw picregq
56%define sixtap_filter_hb picregq56%define sixtap_filter_hb picregq
57%define sixtap_filter_v picregq57%define sixtap_filter_v picregq
@@ -84,7 +84,7 @@ SECTION .text · 26 unmodified lines
84%if WIN6484%if WIN64
85 movsxd %1q, %1d85 movsxd %1q, %1d
86%endif86%endif
87%ifdef PIC87%if PIC
88 add %1q, picregq88 add %1q, picregq
89%else89%else
90 add %1q, %290 add %1q, %2
@@ -104,7 +104,7 @@ SECTION .text · 13 unmodified lines
104 104
105%macro FILTER_V 1105%macro FILTER_V 1
106cglobal %1_rv40_qpel_v, 6,6+npicregs,12, dst, dststride, src, srcstride, height, my, picreg106cglobal %1_rv40_qpel_v, 6,6+npicregs,12, dst, dststride, src, srcstride, height, my, picreg
107%ifdef PIC107%if PIC
108 lea picregq, [sixtap_filter_v_m]108 lea picregq, [sixtap_filter_v_m]
109%endif109%endif
110 pxor m7, m7110 pxor m7, m7
@@ -175,7 +175,7 @@ cglobal %1_rv40_qpel_v, 6,6+npicregs,12, dst, dststride, src, srcstride, height, · 64 unmodified lines
175 175
176%macro FILTER_H 1176%macro FILTER_H 1
177cglobal %1_rv40_qpel_h, 6, 6+npicregs, 12, dst, dststride, src, srcstride, height, mx, picreg177cglobal %1_rv40_qpel_h, 6, 6+npicregs, 12, dst, dststride, src, srcstride, height, mx, picreg
178%ifdef PIC178%if PIC
179 lea picregq, [sixtap_filter_v_m]179 lea picregq, [sixtap_filter_v_m]
180%endif180%endif
181 pxor m7, m7181 pxor m7, m7
@@ -238,7 +238,7 @@ FILTER_V avg · 56 unmodified lines
238 238
239%macro FILTER_SSSE3 1239%macro FILTER_SSSE3 1
240cglobal %1_rv40_qpel_v, 6,6+npicregs,8, dst, dststride, src, srcstride, height, my, picreg240cglobal %1_rv40_qpel_v, 6,6+npicregs,8, dst, dststride, src, srcstride, height, my, picreg
241%ifdef PIC241%if PIC
242 lea picregq, [sixtap_filter_hb_m]242 lea picregq, [sixtap_filter_hb_m]
243%endif243%endif
244 244
@@ -283,7 +283,7 @@ cglobal %1_rv40_qpel_v, 6,6+npicregs,8, dst, dststride, src, srcstride, height, · 38 unmodified lines
283 RET283 RET
284 284
285cglobal %1_rv40_qpel_h, 6,6+npicregs,8, dst, dststride, src, srcstride, height, mx, picreg285cglobal %1_rv40_qpel_h, 6,6+npicregs,8, dst, dststride, src, srcstride, height, mx, picreg
286%ifdef PIC286%if PIC
287 lea picregq, [sixtap_filter_hb_m]287 lea picregq, [sixtap_filter_hb_m]
288%endif288%endif
289 mova m3, [filter_h6_shuf2]289 mova m3, [filter_h6_shuf2]
libavcodec/x86/sbrdsp.asm+3−3
@@ -308,7 +308,7 @@ cglobal sbr_qmf_pre_shuffle, 1,4,6,z · 307 unmodified lines
308 movq [r2q], m2308 movq [r2q], m2
309 RET309 RET
310 310
311%ifdef PIC311%if PIC
312%define NREGS 1312%define NREGS 1
313%if UNIX64313%if UNIX64
314%define NOISE_TABLE r6q ; r5q is m_max314%define NOISE_TABLE r6q ; r5q is m_max
@@ -321,7 +321,7 @@ cglobal sbr_qmf_pre_shuffle, 1,4,6,z · 6 unmodified lines
321%endif321%endif
322 322
323%macro LOAD_NST 1323%macro LOAD_NST 1
324%ifdef PIC324%if PIC
325 lea NOISE_TABLE, [%1]325 lea NOISE_TABLE, [%1]
326 mova m0, [kxq + NOISE_TABLE]326 mova m0, [kxq + NOISE_TABLE]
327%else327%else
@@ -371,7 +371,7 @@ apply_noise_main: · 43 unmodified lines
371 movsxdifnidn noiseq, noised371 movsxdifnidn noiseq, noised
372 dec noiseq372 dec noiseq
373 shl countd, 2373 shl countd, 2
374%ifdef PIC374%if PIC
375 lea NOISE_TABLE, [sbr_noise_table]375 lea NOISE_TABLE, [sbr_noise_table]
376%endif376%endif
377 lea Yq, [Yq + 2*countq]377 lea Yq, [Yq + 2*countq]
libavcodec/x86/vp8dsp.asm+15−15
@@ -114,7 +114,7 @@ bilinear_filter_vb_m: times 8 db 7, 1 · 113 unmodified lines
114 times 8 db 2, 6114 times 8 db 2, 6
115 times 8 db 1, 7115 times 8 db 1, 7
116 116
117%ifdef PIC117%if PIC
118%define fourtap_filter_hw picregq118%define fourtap_filter_hw picregq
119%define sixtap_filter_hw picregq119%define sixtap_filter_hw picregq
120%define fourtap_filter_hb picregq120%define fourtap_filter_hb picregq
@@ -166,7 +166,7 @@ cglobal put_vp8_epel%1_h6, 6, 6 + npicregs, 8, dst, dststride, src, srcstride, h · 45 unmodified lines
166 lea mxd, [mxq*3]166 lea mxd, [mxq*3]
167 mova m3, [filter_h6_shuf2]167 mova m3, [filter_h6_shuf2]
168 mova m4, [filter_h6_shuf3]168 mova m4, [filter_h6_shuf3]
169%ifdef PIC169%if PIC
170 lea picregq, [sixtap_filter_hb_m]170 lea picregq, [sixtap_filter_hb_m]
171%endif171%endif
172 mova m5, [sixtap_filter_hb+mxq*8-48] ; set up 6tap filter in bytes172 mova m5, [sixtap_filter_hb+mxq*8-48] ; set up 6tap filter in bytes
@@ -207,7 +207,7 @@ cglobal put_vp8_epel%1_h4, 6, 6 + npicregs, 7, dst, dststride, src, srcstride, h · 34 unmodified lines
207 mova m2, [pw_256]207 mova m2, [pw_256]
208 mova m3, [filter_h2_shuf]208 mova m3, [filter_h2_shuf]
209 mova m4, [filter_h4_shuf]209 mova m4, [filter_h4_shuf]
210%ifdef PIC210%if PIC
211 lea picregq, [fourtap_filter_hb_m]211 lea picregq, [fourtap_filter_hb_m]
212%endif212%endif
213 mova m5, [fourtap_filter_hb+mxq-16] ; set up 4tap filter in bytes213 mova m5, [fourtap_filter_hb+mxq-16] ; set up 4tap filter in bytes
@@ -234,7 +234,7 @@ cglobal put_vp8_epel%1_h4, 6, 6 + npicregs, 7, dst, dststride, src, srcstride, h · 20 unmodified lines
234 234
235cglobal put_vp8_epel%1_v4, 7, 7, 8, dst, dststride, src, srcstride, height, picreg, my235cglobal put_vp8_epel%1_v4, 7, 7, 8, dst, dststride, src, srcstride, height, picreg, my
236 shl myd, 4236 shl myd, 4
237%ifdef PIC237%if PIC
238 lea picregq, [fourtap_filter_hb_m]238 lea picregq, [fourtap_filter_hb_m]
239%endif239%endif
240 mova m5, [fourtap_filter_hb+myq-16]240 mova m5, [fourtap_filter_hb+myq-16]
@@ -272,7 +272,7 @@ cglobal put_vp8_epel%1_v4, 7, 7, 8, dst, dststride, src, srcstride, height, picr · 31 unmodified lines
272 272
273cglobal put_vp8_epel%1_v6, 7, 7, 8, dst, dststride, src, srcstride, height, picreg, my273cglobal put_vp8_epel%1_v6, 7, 7, 8, dst, dststride, src, srcstride, height, picreg, my
274 lea myd, [myq*3]274 lea myd, [myq*3]
275%ifdef PIC275%if PIC
276 lea picregq, [sixtap_filter_hb_m]276 lea picregq, [sixtap_filter_hb_m]
277%endif277%endif
278 lea myq, [sixtap_filter_hb+myq*8]278 lea myq, [sixtap_filter_hb+myq*8]
@@ -326,7 +326,7 @@ FILTER_SSSE3 8 · 47 unmodified lines
326INIT_MMX mmxext326INIT_MMX mmxext
327cglobal put_vp8_epel4_h4, 6, 6 + npicregs, 0, dst, dststride, src, srcstride, height, mx, picreg327cglobal put_vp8_epel4_h4, 6, 6 + npicregs, 0, dst, dststride, src, srcstride, height, mx, picreg
328 shl mxd, 4328 shl mxd, 4
329%ifdef PIC329%if PIC
330 lea picregq, [fourtap_filter_hw_m]330 lea picregq, [fourtap_filter_hw_m]
331%endif331%endif
332 movq mm4, [fourtap_filter_hw+mxq-16] ; set up 4tap filter in words332 movq mm4, [fourtap_filter_hw+mxq-16] ; set up 4tap filter in words
@@ -374,7 +374,7 @@ cglobal put_vp8_epel4_h4, 6, 6 + npicregs, 0, dst, dststride, src, srcstride, he · 41 unmodified lines
374INIT_MMX mmxext374INIT_MMX mmxext
375cglobal put_vp8_epel4_h6, 6, 6 + npicregs, 0, dst, dststride, src, srcstride, height, mx, picreg375cglobal put_vp8_epel4_h6, 6, 6 + npicregs, 0, dst, dststride, src, srcstride, height, mx, picreg
376 lea mxd, [mxq*3]376 lea mxd, [mxq*3]
377%ifdef PIC377%if PIC
378 lea picregq, [sixtap_filter_hw_m]378 lea picregq, [sixtap_filter_hw_m]
379%endif379%endif
380 movq mm4, [sixtap_filter_hw+mxq*8-48] ; set up 4tap filter in words380 movq mm4, [sixtap_filter_hw+mxq*8-48] ; set up 4tap filter in words
@@ -431,7 +431,7 @@ cglobal put_vp8_epel4_h6, 6, 6 + npicregs, 0, dst, dststride, src, srcstride, he · 50 unmodified lines
431INIT_XMM sse2431INIT_XMM sse2
432cglobal put_vp8_epel8_h4, 6, 6 + npicregs, 10, dst, dststride, src, srcstride, height, mx, picreg432cglobal put_vp8_epel8_h4, 6, 6 + npicregs, 10, dst, dststride, src, srcstride, height, mx, picreg
433 shl mxd, 5433 shl mxd, 5
434%ifdef PIC434%if PIC
435 lea picregq, [fourtap_filter_v_m]435 lea picregq, [fourtap_filter_v_m]
436%endif436%endif
437 lea mxq, [fourtap_filter_v+mxq-32]437 lea mxq, [fourtap_filter_v+mxq-32]
@@ -480,7 +480,7 @@ INIT_XMM sse2 · 42 unmodified lines
480cglobal put_vp8_epel8_h6, 6, 6 + npicregs, 14, dst, dststride, src, srcstride, height, mx, picreg480cglobal put_vp8_epel8_h6, 6, 6 + npicregs, 14, dst, dststride, src, srcstride, height, mx, picreg
481 lea mxd, [mxq*3]481 lea mxd, [mxq*3]
482 shl mxd, 4482 shl mxd, 4
483%ifdef PIC483%if PIC
484 lea picregq, [sixtap_filter_v_m]484 lea picregq, [sixtap_filter_v_m]
485%endif485%endif
486 lea mxq, [sixtap_filter_v+mxq-96]486 lea mxq, [sixtap_filter_v+mxq-96]
@@ -543,7 +543,7 @@ cglobal put_vp8_epel8_h6, 6, 6 + npicregs, 14, dst, dststride, src, srcstride, h · 56 unmodified lines
543; 4x4 block, V-only 4-tap filter543; 4x4 block, V-only 4-tap filter
544cglobal put_vp8_epel%1_v4, 7, 7, 8, dst, dststride, src, srcstride, height, picreg, my544cglobal put_vp8_epel%1_v4, 7, 7, 8, dst, dststride, src, srcstride, height, picreg, my
545 shl myd, 5545 shl myd, 5
546%ifdef PIC546%if PIC
547 lea picregq, [fourtap_filter_v_m]547 lea picregq, [fourtap_filter_v_m]
548%endif548%endif
549 lea myq, [fourtap_filter_v+myq-32]549 lea myq, [fourtap_filter_v+myq-32]
@@ -597,7 +597,7 @@ cglobal put_vp8_epel%1_v4, 7, 7, 8, dst, dststride, src, srcstride, height, picr · 47 unmodified lines
597cglobal put_vp8_epel%1_v6, 7, 7, 8, dst, dststride, src, srcstride, height, picreg, my597cglobal put_vp8_epel%1_v6, 7, 7, 8, dst, dststride, src, srcstride, height, picreg, my
598 shl myd, 4598 shl myd, 4
599 lea myq, [myq*3]599 lea myq, [myq*3]
600%ifdef PIC600%if PIC
601 lea picregq, [sixtap_filter_v_m]601 lea picregq, [sixtap_filter_v_m]
602%endif602%endif
603 lea myq, [sixtap_filter_v+myq-96]603 lea myq, [sixtap_filter_v+myq-96]
@@ -667,7 +667,7 @@ FILTER_V 8 · 63 unmodified lines
667%if cpuflag(ssse3)667%if cpuflag(ssse3)
668cglobal put_vp8_bilinear%1_v, 7, 7, 5, dst, dststride, src, srcstride, height, picreg, my668cglobal put_vp8_bilinear%1_v, 7, 7, 5, dst, dststride, src, srcstride, height, picreg, my
669 shl myd, 4669 shl myd, 4
670%ifdef PIC670%if PIC
671 lea picregq, [bilinear_filter_vb_m]671 lea picregq, [bilinear_filter_vb_m]
672%endif672%endif
673 pxor m4, m4673 pxor m4, m4
@@ -697,7 +697,7 @@ cglobal put_vp8_bilinear%1_v, 7, 7, 5, dst, dststride, src, srcstride, height, p · 23 unmodified lines
697%else ; cpuflag(ssse3)697%else ; cpuflag(ssse3)
698cglobal put_vp8_bilinear%1_v, 7, 7, 7, dst, dststride, src, srcstride, height, picreg, my698cglobal put_vp8_bilinear%1_v, 7, 7, 7, dst, dststride, src, srcstride, height, picreg, my
699 shl myd, 4699 shl myd, 4
700%ifdef PIC700%if PIC
701 lea picregq, [bilinear_filter_vw_m]701 lea picregq, [bilinear_filter_vw_m]
702%endif702%endif
703 pxor m6, m6703 pxor m6, m6
@@ -743,7 +743,7 @@ cglobal put_vp8_bilinear%1_v, 7, 7, 7, dst, dststride, src, srcstride, height, p · 39 unmodified lines
743%if cpuflag(ssse3)743%if cpuflag(ssse3)
744cglobal put_vp8_bilinear%1_h, 6, 6 + npicregs, 5, dst, dststride, src, srcstride, height, mx, picreg744cglobal put_vp8_bilinear%1_h, 6, 6 + npicregs, 5, dst, dststride, src, srcstride, height, mx, picreg
745 shl mxd, 4745 shl mxd, 4
746%ifdef PIC746%if PIC
747 lea picregq, [bilinear_filter_vb_m]747 lea picregq, [bilinear_filter_vb_m]
748%endif748%endif
749 pxor m4, m4749 pxor m4, m4
@@ -773,7 +773,7 @@ cglobal put_vp8_bilinear%1_h, 6, 6 + npicregs, 5, dst, dststride, src, srcstride · 23 unmodified lines
773%else ; cpuflag(ssse3)773%else ; cpuflag(ssse3)
774cglobal put_vp8_bilinear%1_h, 6, 6 + npicregs, 7, dst, dststride, src, srcstride, height, mx, picreg774cglobal put_vp8_bilinear%1_h, 6, 6 + npicregs, 7, dst, dststride, src, srcstride, height, mx, picreg
775 shl mxd, 4775 shl mxd, 4
776%ifdef PIC776%if PIC
777 lea picregq, [bilinear_filter_vw_m]777 lea picregq, [bilinear_filter_vw_m]
778%endif778%endif
779 pxor m6, m6779 pxor m6, m6
libavcodec/x86/vp9itxfm.asm+2−0
@@ -330,7 +330,9 @@ IDCT_4x4_FN ssse3 · 329 unmodified lines
330INIT_MMX %5330INIT_MMX %5
331cglobal vp9_%1_%3_4x4_add, 3, 3, 0, dst, stride, block, eob331cglobal vp9_%1_%3_4x4_add, 3, 3, 0, dst, stride, block, eob
332%if WIN64 && notcpuflag(ssse3)332%if WIN64 && notcpuflag(ssse3)
333INIT_XMM cpuname
333 WIN64_SPILL_XMM 8334 WIN64_SPILL_XMM 8
335INIT_MMX cpuname
334%endif336%endif
335 movdqa xmm5, [pd_8192]337 movdqa xmm5, [pd_8192]
336 mova m0, [blockq+ 0]338 mova m0, [blockq+ 0]
libavcodec/x86/vp9itxfm_16bpp.asm+7−5
@@ -303,7 +303,9 @@ IDCT4_10_FN · 302 unmodified lines
303%macro IADST4_FN 4303%macro IADST4_FN 4
304cglobal vp9_%1_%3_4x4_add_10, 3, 3, 0, dst, stride, block, eob304cglobal vp9_%1_%3_4x4_add_10, 3, 3, 0, dst, stride, block, eob
305%if WIN64 && notcpuflag(ssse3)305%if WIN64 && notcpuflag(ssse3)
306INIT_XMM cpuname
306 WIN64_SPILL_XMM 8307 WIN64_SPILL_XMM 8
308INIT_MMX cpuname
307%endif309%endif
308 movdqa xmm5, [pd_8192]310 movdqa xmm5, [pd_8192]
309 mova m0, [blockq+0*16+0]311 mova m0, [blockq+0*16+0]
@@ -672,7 +674,7 @@ cglobal vp9_idct_idct_8x8_add_10, 4, 6 + ARCH_X86_64, 14, \ · 362 unmodified lines
672 mov dstbakq, dstq674 mov dstbakq, dstq
673 movsxd cntq, cntd675 movsxd cntq, cntd
674%endif676%endif
675%ifdef PIC677%if PIC
676 lea ptrq, [default_8x8]678 lea ptrq, [default_8x8]
677 movzx cntd, byte [ptrq+cntq-1]679 movzx cntd, byte [ptrq+cntq-1]
678%else680%else
@@ -921,7 +923,7 @@ cglobal vp9_%1_%3_8x8_add_10, 4, 6 + ARCH_X86_64, 16, \ · 242 unmodified lines
921 mov dstbakq, dstq923 mov dstbakq, dstq
922 movsxd cntq, cntd924 movsxd cntq, cntd
923%endif925%endif
924%ifdef PIC926%if PIC
925 lea ptrq, [%5_8x8]927 lea ptrq, [%5_8x8]
926 movzx cntd, byte [ptrq+cntq-1]928 movzx cntd, byte [ptrq+cntq-1]
927%else929%else
@@ -1128,7 +1130,7 @@ cglobal vp9_idct_idct_16x16_add_10, 4, 6 + ARCH_X86_64, 16, \ · 200 unmodified lines
1128 mov dstbakq, dstq1130 mov dstbakq, dstq
1129 movsxd cntq, cntd1131 movsxd cntq, cntd
1130%endif1132%endif
1131%ifdef PIC1133%if PIC
1132 lea ptrq, [default_16x16]1134 lea ptrq, [default_16x16]
1133 movzx cntd, byte [ptrq+cntq-1]1135 movzx cntd, byte [ptrq+cntq-1]
1134%else1136%else
@@ -1445,7 +1447,7 @@ cglobal vp9_%1_%4_16x16_add_10, 4, 6 + ARCH_X86_64, 16, \ · 310 unmodified lines
1445 mov dstbakq, dstq1447 mov dstbakq, dstq
1446 movsxd cntq, cntd1448 movsxd cntq, cntd
1447%endif1449%endif
1448%ifdef PIC1450%if PIC
1449 lea ptrq, [%7_16x16]1451 lea ptrq, [%7_16x16]
1450 movzx cntd, byte [ptrq+cntq-1]1452 movzx cntd, byte [ptrq+cntq-1]
1451%else1453%else
@@ -1958,7 +1960,7 @@ cglobal vp9_idct_idct_32x32_add_10, 4, 6 + ARCH_X86_64, 16, \ · 506 unmodified lines
1958 mov dstbakq, dstq1960 mov dstbakq, dstq
1959 movsxd cntq, cntd1961 movsxd cntq, cntd
1960%endif1962%endif
1961%ifdef PIC1963%if PIC
1962 lea ptrq, [default_32x32]1964 lea ptrq, [default_32x32]
1963 movzx cntd, byte [ptrq+cntq-1]1965 movzx cntd, byte [ptrq+cntq-1]
1964%else1966%else

Keyboard shortcuts

General

Show keyboard shortcuts
?
Focus the current search
/

Issues and pull requests

Open filters
f

Issues, pull requests and Inbox

Select next item
j↓
Select previous item
k↑
Open selected item
Enter
Dismiss filters or clear selection
Esc

Repository code

Go to file
t