sws/aarch64: add ff_yuv2planeX_8_neon
authorClément Bœsch <clement@stupeflix.com>
Fri, 1 Apr 2016 15:27:29 +0000 (17:27 +0200)
committerClément Bœsch <clement@stupeflix.com>
Mon, 11 Apr 2016 14:27:19 +0000 (16:27 +0200)
libswscale/aarch64/Makefile
libswscale/aarch64/output.S [new file with mode: 0644]
libswscale/aarch64/swscale.c
libswscale/utils.c

index 51bff08..64a3fe2 100644 (file)
@@ -2,4 +2,5 @@ OBJS        += aarch64/swscale.o                \
                aarch64/swscale_unscaled.o       \
 
 NEON-OBJS   += aarch64/hscale.o                 \
+               aarch64/output.o                 \
                aarch64/yuv2rgb_neon.o           \
diff --git a/libswscale/aarch64/output.S b/libswscale/aarch64/output.S
new file mode 100644 (file)
index 0000000..90d3b57
--- /dev/null
@@ -0,0 +1,66 @@
+/*
+ * Copyright (c) 2016 Clément Bœsch <clement stupeflix.com>
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#include "libavutil/aarch64/asm.S"
+
+function ff_yuv2planeX_8_neon, export=1
+        ld1                 {v0.8B}, [x5]                   // load 8x8-bit dither
+        cbz                 w6, 1f                          // check if offsetting present
+        ext                 v0.8B, v0.8B, v0.8B, #3         // honor offsetting which can be 0 or 3 only
+1:      uxtl                v0.8H, v0.8B                    // extend dither to 16-bit
+        ushll               v1.4S, v0.4H, #12               // extend dither to 32-bit with left shift by 12 (part 1)
+        ushll2              v2.4S, v0.8H, #12               // extend dither to 32-bit with left shift by 12 (part 2)
+        mov                 x7, #0                          // i = 0
+2:      mov                 v3.16B, v1.16B                  // initialize accumulator part 1 with dithering value
+        mov                 v4.16B, v2.16B                  // initialize accumulator part 2 with dithering value
+        mov                 w8, w1                          // tmpfilterSize = filterSize
+        mov                 x9, x2                          // srcp    = src
+        mov                 x10, x0                         // filterp = filter
+3:      ldp                 x11, x12, [x9], #16             // get 2 pointers: src[j] and src[j+1]
+        add                 x11, x11, x7, lsl #1            // &src[j  ][i]
+        add                 x12, x12, x7, lsl #1            // &src[j+1][i]
+        ld1                 {v5.8H}, [x11]                  // read 8x16-bit @ src[j  ][i + {0..7}]: A,B,C,D,E,F,G,H
+        ld1                 {v6.8H}, [x12]                  // read 8x16-bit @ src[j+1][i + {0..7}]: I,J,K,L,M,N,O,P
+        ldr                 w11, [x10], #4                  // read 2x16-bit coeffs (X, Y) at (filter[j], filter[j+1])
+        zip1                v16.8H, v5.8H, v6.8H            // A,I,B,J,C,K,D,L
+        zip2                v17.8H, v5.8H, v6.8H            // E,M,F,N,F,O,H,P
+        dup                 v7.4S, w11                      // X,Y,X,Y,X,Y,X,Y
+        smull               v18.4S, v16.4H, v7.4H           // A.X I.Y B.X J.Y
+        smull               v20.4S, v17.4H, v7.4H           // E.X M.Y F.X N.Y
+        smull2              v19.4S, v16.8H, v7.8H           // C.X K.Y D.X L.Y
+        smull2              v21.4S, v17.8H, v7.8H           // G.X O.Y H.X P.Y
+        addp                v16.4S, v18.4S, v19.4S          // A.X+I.Y B.X+J.Y C.X+K.Y D.X+L.Y
+        addp                v17.4S, v20.4S, v21.4S          // E.X+M.Y F.X+N.Y F.X+O.Y H.X+P.Y
+        add                 v3.4S, v3.4S, v16.4S            // update val accumulator for part 1
+        add                 v4.4S, v4.4S, v17.4S            // update val accumulator for part 2
+        subs                w8, w8, #2                      // tmpfilterSize -= 2
+        b.gt                3b                              // loop until filterSize consumed
+        sshr                v3.4S, v3.4S, #19               // val>>19 (part 1)
+        sshr                v4.4S, v4.4S, #19               // val>>19 (part 2)
+        sqxtun              v3.4H, v3.4S                    // clip16(val>>19) (part 1)
+        sqxtun              v4.4H, v4.4S                    // clip16(val>>19) (part 2)
+        mov                 v3.D[1], v4.D[0]                // merge part 1 and part 2
+        uqxtn               v3.8B, v3.8H                    // clip8(val>>19)
+        st1                 {v3.1D}, [x3], #8               // write to destination
+        add                 x7, x7, #8                      // i += 8
+        subs                w4, w4, #8                      // dstW -= 8
+        b.gt                2b                              // loop until width consumed
+        ret
+endfunc
index 6b9b046..54a3bea 100644 (file)
@@ -25,6 +25,10 @@ void ff_hscale_8_to_15_neon(SwsContext *c, int16_t *dst, int dstW,
                             const uint8_t *src, const int16_t *filter,
                             const int32_t *filterPos, int filterSize);
 
+void ff_yuv2planeX_8_neon(const int16_t *filter, int filterSize,
+                          const int16_t **src, uint8_t *dest, int dstW,
+                          const uint8_t *dither, int offset);
+
 av_cold void ff_sws_init_swscale_aarch64(SwsContext *c)
 {
     int cpu_flags = av_get_cpu_flags();
@@ -33,5 +37,8 @@ av_cold void ff_sws_init_swscale_aarch64(SwsContext *c)
         if (c->srcBpc == 8 && c->dstBpc <= 14) {
             c->hyScale = c->hcScale = ff_hscale_8_to_15_neon;
         }
+        if (c->dstBpc == 8) {
+            c->yuv2planeX = ff_yuv2planeX_8_neon;
+        }
     }
 }
index b81661f..63a8226 100644 (file)
@@ -1655,7 +1655,8 @@ av_cold int sws_init_context(SwsContext *c, SwsFilter *srcFilter,
     /* precalculate vertical scaler filter coefficients */
     {
         const int filterAlign = X86_MMX(cpu_flags)     ? 2 :
-                                PPC_ALTIVEC(cpu_flags) ? 8 : 1;
+                                PPC_ALTIVEC(cpu_flags) ? 8 :
+                                have_neon(cpu_flags)   ? 2 : 1;
 
         if ((ret = initFilter(&c->vLumFilter, &c->vLumFilterPos, &c->vLumFilterSize,
                        c->lumYInc, srcH, dstH, filterAlign, (1 << 12),