Merge remote-tracking branch 'qatar/master'
authorMichael Niedermayer <michaelni@gmx.at>
Fri, 14 Sep 2012 12:07:21 +0000 (14:07 +0200)
committerMichael Niedermayer <michaelni@gmx.at>
Fri, 14 Sep 2012 12:13:38 +0000 (14:13 +0200)
* qatar/master:
  x86: dsputil: Move Xvid IDCT put/add functions to a more suitable place
  trasher: Include all the necessary headers
  x86: Remove some leftover declarations for non-existent functions
  ARM: libavresample: NEON optimised generic fltp to s16 conversion
  ARM: libavresample: NEON optimised stereo fltp to s16 conversion
  ARM: libavresample: NEON optimised flat float to s16 conversion

Conflicts:
libavcodec/x86/dsputil_mmx.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>
libavcodec/dct-test.c
libavcodec/x86/dsputil_mmx.c
libavcodec/x86/idct_mmx_xvid.c
libavcodec/x86/idct_xvid.h
libavresample/arm/Makefile [new file with mode: 0644]
libavresample/arm/audio_convert_init.c [new file with mode: 0644]
libavresample/arm/audio_convert_neon.S [new file with mode: 0644]
libavresample/audio_convert.c
libavresample/audio_convert.h
tools/trasher.c

index a277817..6cefd9d 100644 (file)
@@ -57,7 +57,6 @@ void ff_bfin_fdct(DCTELEM *block);
 
 // ALTIVEC
 void ff_fdct_altivec(DCTELEM *block);
-//void ff_idct_altivec(DCTELEM *block);?? no routine
 
 // ARM
 void ff_j_rev_dct_arm(DCTELEM *data);
index 567d8be..c36a027 100644 (file)
@@ -2291,30 +2291,6 @@ static void ff_libmpeg2mmx2_idct_add(uint8_t *dest, int line_size,
 }
 #endif
 
-static void ff_idct_xvid_mmx_put(uint8_t *dest, int line_size, DCTELEM *block)
-{
-    ff_idct_xvid_mmx(block);
-    ff_put_pixels_clamped_mmx(block, dest, line_size);
-}
-
-static void ff_idct_xvid_mmx_add(uint8_t *dest, int line_size, DCTELEM *block)
-{
-    ff_idct_xvid_mmx(block);
-    ff_add_pixels_clamped_mmx(block, dest, line_size);
-}
-
-static void ff_idct_xvid_mmx2_put(uint8_t *dest, int line_size, DCTELEM *block)
-{
-    ff_idct_xvid_mmx2(block);
-    ff_put_pixels_clamped_mmx(block, dest, line_size);
-}
-
-static void ff_idct_xvid_mmx2_add(uint8_t *dest, int line_size, DCTELEM *block)
-{
-    ff_idct_xvid_mmx2(block);
-    ff_add_pixels_clamped_mmx(block, dest, line_size);
-}
-
 static void vorbis_inverse_coupling_3dnow(float *mag, float *ang, int blocksize)
 {
     int i;
index 29286b6..2465cdf 100644 (file)
@@ -44,6 +44,7 @@
 #include "config.h"
 #include "libavcodec/avcodec.h"
 #include "libavutil/mem.h"
+#include "dsputil_mmx.h"
 #include "idct_xvid.h"
 
 #if HAVE_INLINE_ASM
@@ -529,4 +530,28 @@ __asm__ volatile(
     :: "r"(block), "r"(rounder_0), "r"(tab_i_04_xmm), "r"(tg_1_16));
 }
 
+void ff_idct_xvid_mmx_put(uint8_t *dest, int line_size, DCTELEM *block)
+{
+    ff_idct_xvid_mmx(block);
+    ff_put_pixels_clamped_mmx(block, dest, line_size);
+}
+
+void ff_idct_xvid_mmx_add(uint8_t *dest, int line_size, DCTELEM *block)
+{
+    ff_idct_xvid_mmx(block);
+    ff_add_pixels_clamped_mmx(block, dest, line_size);
+}
+
+void ff_idct_xvid_mmx2_put(uint8_t *dest, int line_size, DCTELEM *block)
+{
+    ff_idct_xvid_mmx2(block);
+    ff_put_pixels_clamped_mmx(block, dest, line_size);
+}
+
+void ff_idct_xvid_mmx2_add(uint8_t *dest, int line_size, DCTELEM *block)
+{
+    ff_idct_xvid_mmx2(block);
+    ff_add_pixels_clamped_mmx(block, dest, line_size);
+}
+
 #endif /* HAVE_INLINE_ASM */
index be91d1c..6000654 100644 (file)
 
 #include <stdint.h>
 
+#include "libavcodec/dsputil.h"
+
 void ff_idct_xvid_mmx(short *block);
+void ff_idct_xvid_mmx_put(uint8_t *dest, int line_size, DCTELEM *block);
+void ff_idct_xvid_mmx_add(uint8_t *dest, int line_size, DCTELEM *block);
+
 void ff_idct_xvid_mmx2(short *block);
+void ff_idct_xvid_mmx2_put(uint8_t *dest, int line_size, DCTELEM *block);
+void ff_idct_xvid_mmx2_add(uint8_t *dest, int line_size, DCTELEM *block);
+
 void ff_idct_xvid_sse2(short *block);
 void ff_idct_xvid_sse2_put(uint8_t *dest, int line_size, short *block);
 void ff_idct_xvid_sse2_add(uint8_t *dest, int line_size, short *block);
diff --git a/libavresample/arm/Makefile b/libavresample/arm/Makefile
new file mode 100644 (file)
index 0000000..55683cb
--- /dev/null
@@ -0,0 +1,2 @@
+OBJS      += arm/audio_convert_init.o
+NEON-OBJS += arm/audio_convert_neon.o
diff --git a/libavresample/arm/audio_convert_init.c b/libavresample/arm/audio_convert_init.c
new file mode 100644 (file)
index 0000000..3d19a0e
--- /dev/null
@@ -0,0 +1,49 @@
+/*
+ * 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 <stdint.h>
+
+#include "config.h"
+#include "libavutil/attributes.h"
+#include "libavutil/cpu.h"
+#include "libavutil/arm/cpu.h"
+#include "libavutil/samplefmt.h"
+#include "libavresample/audio_convert.h"
+
+void ff_conv_flt_to_s16_neon(int16_t *dst, const float *src, int len);
+void ff_conv_fltp_to_s16_neon(int16_t *dst, float *const *src,
+                              int len, int channels);
+void ff_conv_fltp_to_s16_2ch_neon(int16_t *dst, float *const *src,
+                                  int len, int channels);
+
+av_cold void ff_audio_convert_init_arm(AudioConvert *ac)
+{
+    int cpu_flags = av_get_cpu_flags();
+
+    if (have_neon(cpu_flags)) {
+        ff_audio_convert_set_func(ac, AV_SAMPLE_FMT_S16, AV_SAMPLE_FMT_FLT,
+                                  0, 16, 8, "NEON",
+                                  ff_conv_flt_to_s16_neon);
+        ff_audio_convert_set_func(ac, AV_SAMPLE_FMT_S16, AV_SAMPLE_FMT_FLTP,
+                                  0, 16, 8, "NEON",
+                                  ff_conv_fltp_to_s16_neon);
+        ff_audio_convert_set_func(ac, AV_SAMPLE_FMT_S16, AV_SAMPLE_FMT_FLTP,
+                                  2, 16, 8, "NEON",
+                                  ff_conv_fltp_to_s16_2ch_neon);
+    }
+}
diff --git a/libavresample/arm/audio_convert_neon.S b/libavresample/arm/audio_convert_neon.S
new file mode 100644 (file)
index 0000000..98f77f0
--- /dev/null
@@ -0,0 +1,363 @@
+/*
+ * Copyright (c) 2008 Mans Rullgard <mans@mansr.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 "config.h"
+#include "libavutil/arm/asm.S"
+
+function ff_conv_flt_to_s16_neon, export=1
+        subs            r2,  r2,  #8
+        vld1.32         {q0},     [r1,:128]!
+        vcvt.s32.f32    q8,  q0,  #31
+        vld1.32         {q1},     [r1,:128]!
+        vcvt.s32.f32    q9,  q1,  #31
+        beq             3f
+        bics            r12, r2,  #15
+        beq             2f
+1:      subs            r12, r12, #16
+        vqrshrn.s32     d4,  q8,  #16
+        vld1.32         {q0},     [r1,:128]!
+        vcvt.s32.f32    q0,  q0,  #31
+        vqrshrn.s32     d5,  q9,  #16
+        vld1.32         {q1},     [r1,:128]!
+        vcvt.s32.f32    q1,  q1,  #31
+        vqrshrn.s32     d6,  q0,  #16
+        vst1.16         {q2},     [r0,:128]!
+        vqrshrn.s32     d7,  q1,  #16
+        vld1.32         {q8},     [r1,:128]!
+        vcvt.s32.f32    q8,  q8,  #31
+        vld1.32         {q9},     [r1,:128]!
+        vcvt.s32.f32    q9,  q9,  #31
+        vst1.16         {q3},     [r0,:128]!
+        bne             1b
+        ands            r2,  r2,  #15
+        beq             3f
+2:      vld1.32         {q0},     [r1,:128]!
+        vqrshrn.s32     d4,  q8,  #16
+        vcvt.s32.f32    q0,  q0,  #31
+        vld1.32         {q1},     [r1,:128]!
+        vqrshrn.s32     d5,  q9,  #16
+        vcvt.s32.f32    q1,  q1,  #31
+        vqrshrn.s32     d6,  q0,  #16
+        vst1.16         {q2},     [r0,:128]!
+        vqrshrn.s32     d7,  q1,  #16
+        vst1.16         {q3},     [r0,:128]!
+        bx              lr
+3:      vqrshrn.s32     d4,  q8,  #16
+        vqrshrn.s32     d5,  q9,  #16
+        vst1.16         {q2},     [r0,:128]!
+        bx              lr
+endfunc
+
+function ff_conv_fltp_to_s16_2ch_neon, export=1
+        ldm             r1,  {r1, r3}
+        subs            r2,  r2,  #8
+        vld1.32         {q0},     [r1,:128]!
+        vcvt.s32.f32    q8,  q0,  #31
+        vld1.32         {q1},     [r1,:128]!
+        vcvt.s32.f32    q9,  q1,  #31
+        vld1.32         {q10},    [r3,:128]!
+        vcvt.s32.f32    q10, q10, #31
+        vld1.32         {q11},    [r3,:128]!
+        vcvt.s32.f32    q11, q11, #31
+        beq             3f
+        bics            r12, r2,  #15
+        beq             2f
+1:      subs            r12, r12, #16
+        vld1.32         {q0},     [r1,:128]!
+        vcvt.s32.f32    q0,  q0,  #31
+        vsri.32         q10, q8,  #16
+        vld1.32         {q1},     [r1,:128]!
+        vcvt.s32.f32    q1,  q1,  #31
+        vld1.32         {q12},    [r3,:128]!
+        vcvt.s32.f32    q12, q12, #31
+        vld1.32         {q13},    [r3,:128]!
+        vsri.32         q11, q9,  #16
+        vst1.16         {q10},    [r0,:128]!
+        vcvt.s32.f32    q13, q13, #31
+        vst1.16         {q11},    [r0,:128]!
+        vsri.32         q12, q0,  #16
+        vld1.32         {q8},     [r1,:128]!
+        vsri.32         q13, q1,  #16
+        vst1.16         {q12},    [r0,:128]!
+        vcvt.s32.f32    q8,  q8,  #31
+        vld1.32         {q9},     [r1,:128]!
+        vcvt.s32.f32    q9,  q9,  #31
+        vld1.32         {q10},    [r3,:128]!
+        vcvt.s32.f32    q10, q10, #31
+        vld1.32         {q11},    [r3,:128]!
+        vcvt.s32.f32    q11, q11, #31
+        vst1.16         {q13},    [r0,:128]!
+        bne             1b
+        ands            r2,  r2,  #15
+        beq             3f
+2:      vsri.32         q10, q8,  #16
+        vld1.32         {q0},     [r1,:128]!
+        vcvt.s32.f32    q0,  q0,  #31
+        vld1.32         {q1},     [r1,:128]!
+        vcvt.s32.f32    q1,  q1,  #31
+        vld1.32         {q12},    [r3,:128]!
+        vcvt.s32.f32    q12, q12, #31
+        vsri.32         q11, q9,  #16
+        vld1.32         {q13},    [r3,:128]!
+        vcvt.s32.f32    q13, q13, #31
+        vst1.16         {q10},    [r0,:128]!
+        vsri.32         q12, q0,  #16
+        vst1.16         {q11},    [r0,:128]!
+        vsri.32         q13, q1,  #16
+        vst1.16         {q12-q13},[r0,:128]!
+        bx              lr
+3:      vsri.32         q10, q8,  #16
+        vsri.32         q11, q9,  #16
+        vst1.16         {q10-q11},[r0,:128]!
+        bx              lr
+endfunc
+
+function ff_conv_fltp_to_s16_neon, export=1
+        cmp             r3,  #2
+        itt             lt
+        ldrlt           r1,  [r1]
+        blt             ff_conv_flt_to_s16_neon
+        beq             ff_conv_fltp_to_s16_2ch_neon
+
+        push            {r4-r8, lr}
+        cmp             r3,  #4
+        lsl             r12, r3,  #1
+        blt             4f
+
+        @ 4 channels
+5:      ldm             r1!, {r4-r7}
+        mov             lr,  r2
+        mov             r8,  r0
+        vld1.32         {q8},     [r4,:128]!
+        vcvt.s32.f32    q8,  q8,  #31
+        vld1.32         {q9},     [r5,:128]!
+        vcvt.s32.f32    q9,  q9,  #31
+        vld1.32         {q10},    [r6,:128]!
+        vcvt.s32.f32    q10, q10, #31
+        vld1.32         {q11},    [r7,:128]!
+        vcvt.s32.f32    q11, q11, #31
+6:      subs            lr,  lr,  #8
+        vld1.32         {q0},     [r4,:128]!
+        vcvt.s32.f32    q0,  q0,  #31
+        vsri.32         q9,  q8,  #16
+        vld1.32         {q1},     [r5,:128]!
+        vcvt.s32.f32    q1,  q1,  #31
+        vsri.32         q11, q10, #16
+        vld1.32         {q2},     [r6,:128]!
+        vcvt.s32.f32    q2,  q2,  #31
+        vzip.32         d18, d22
+        vld1.32         {q3},     [r7,:128]!
+        vcvt.s32.f32    q3,  q3,  #31
+        vzip.32         d19, d23
+        vst1.16         {d18},    [r8], r12
+        vsri.32         q1,  q0,  #16
+        vst1.16         {d22},    [r8], r12
+        vsri.32         q3,  q2,  #16
+        vst1.16         {d19},    [r8], r12
+        vzip.32         d2,  d6
+        vst1.16         {d23},    [r8], r12
+        vzip.32         d3,  d7
+        beq             7f
+        vld1.32         {q8},     [r4,:128]!
+        vcvt.s32.f32    q8,  q8,  #31
+        vst1.16         {d2},     [r8], r12
+        vld1.32         {q9},     [r5,:128]!
+        vcvt.s32.f32    q9,  q9,  #31
+        vst1.16         {d6},     [r8], r12
+        vld1.32         {q10},    [r6,:128]!
+        vcvt.s32.f32    q10, q10, #31
+        vst1.16         {d3},     [r8], r12
+        vld1.32         {q11},    [r7,:128]!
+        vcvt.s32.f32    q11, q11, #31
+        vst1.16         {d7},     [r8], r12
+        b               6b
+7:      vst1.16         {d2},     [r8], r12
+        vst1.16         {d6},     [r8], r12
+        vst1.16         {d3},     [r8], r12
+        vst1.16         {d7},     [r8], r12
+        subs            r3,  r3,  #4
+        it              eq
+        popeq           {r4-r8, pc}
+        cmp             r3,  #4
+        add             r0,  r0,  #8
+        bge             5b
+
+        @ 2 channels
+4:      cmp             r3,  #2
+        blt             4f
+        ldm             r1!, {r4-r5}
+        mov             lr,  r2
+        mov             r8,  r0
+        tst             lr,  #8
+        vld1.32         {q8},     [r4,:128]!
+        vcvt.s32.f32    q8,  q8,  #31
+        vld1.32         {q9},     [r5,:128]!
+        vcvt.s32.f32    q9,  q9,  #31
+        vld1.32         {q10},    [r4,:128]!
+        vcvt.s32.f32    q10, q10, #31
+        vld1.32         {q11},    [r5,:128]!
+        vcvt.s32.f32    q11, q11, #31
+        beq             6f
+        subs            lr,  lr,  #8
+        beq             7f
+        vsri.32         d18, d16, #16
+        vsri.32         d19, d17, #16
+        vld1.32         {q8},     [r4,:128]!
+        vcvt.s32.f32    q8,  q8,  #31
+        vst1.32         {d18[0]}, [r8], r12
+        vsri.32         d22, d20, #16
+        vst1.32         {d18[1]}, [r8], r12
+        vsri.32         d23, d21, #16
+        vst1.32         {d19[0]}, [r8], r12
+        vst1.32         {d19[1]}, [r8], r12
+        vld1.32         {q9},     [r5,:128]!
+        vcvt.s32.f32    q9,  q9,  #31
+        vst1.32         {d22[0]}, [r8], r12
+        vst1.32         {d22[1]}, [r8], r12
+        vld1.32         {q10},    [r4,:128]!
+        vcvt.s32.f32    q10, q10, #31
+        vst1.32         {d23[0]}, [r8], r12
+        vst1.32         {d23[1]}, [r8], r12
+        vld1.32         {q11},    [r5,:128]!
+        vcvt.s32.f32    q11, q11, #31
+6:      subs            lr,  lr,  #16
+        vld1.32         {q0},     [r4,:128]!
+        vcvt.s32.f32    q0,  q0,  #31
+        vsri.32         d18, d16, #16
+        vld1.32         {q1},     [r5,:128]!
+        vcvt.s32.f32    q1,  q1,  #31
+        vsri.32         d19, d17, #16
+        vld1.32         {q2},     [r4,:128]!
+        vcvt.s32.f32    q2,  q2,  #31
+        vld1.32         {q3},     [r5,:128]!
+        vcvt.s32.f32    q3,  q3,  #31
+        vst1.32         {d18[0]}, [r8], r12
+        vsri.32         d22, d20, #16
+        vst1.32         {d18[1]}, [r8], r12
+        vsri.32         d23, d21, #16
+        vst1.32         {d19[0]}, [r8], r12
+        vsri.32         d2,  d0,  #16
+        vst1.32         {d19[1]}, [r8], r12
+        vsri.32         d3,  d1,  #16
+        vst1.32         {d22[0]}, [r8], r12
+        vsri.32         d6,  d4,  #16
+        vst1.32         {d22[1]}, [r8], r12
+        vsri.32         d7,  d5,  #16
+        vst1.32         {d23[0]}, [r8], r12
+        vst1.32         {d23[1]}, [r8], r12
+        beq             6f
+        vld1.32         {q8},     [r4,:128]!
+        vcvt.s32.f32    q8,  q8,  #31
+        vst1.32         {d2[0]},  [r8], r12
+        vst1.32         {d2[1]},  [r8], r12
+        vld1.32         {q9},     [r5,:128]!
+        vcvt.s32.f32    q9,  q9,  #31
+        vst1.32         {d3[0]},  [r8], r12
+        vst1.32         {d3[1]},  [r8], r12
+        vld1.32         {q10},    [r4,:128]!
+        vcvt.s32.f32    q10, q10, #31
+        vst1.32         {d6[0]},  [r8], r12
+        vst1.32         {d6[1]},  [r8], r12
+        vld1.32         {q11},    [r5,:128]!
+        vcvt.s32.f32    q11, q11, #31
+        vst1.32         {d7[0]},  [r8], r12
+        vst1.32         {d7[1]},  [r8], r12
+        bgt             6b
+6:      vst1.32         {d2[0]},  [r8], r12
+        vst1.32         {d2[1]},  [r8], r12
+        vst1.32         {d3[0]},  [r8], r12
+        vst1.32         {d3[1]},  [r8], r12
+        vst1.32         {d6[0]},  [r8], r12
+        vst1.32         {d6[1]},  [r8], r12
+        vst1.32         {d7[0]},  [r8], r12
+        vst1.32         {d7[1]},  [r8], r12
+        b               8f
+7:      vsri.32         d18, d16, #16
+        vsri.32         d19, d17, #16
+        vst1.32         {d18[0]}, [r8], r12
+        vsri.32         d22, d20, #16
+        vst1.32         {d18[1]}, [r8], r12
+        vsri.32         d23, d21, #16
+        vst1.32         {d19[0]}, [r8], r12
+        vst1.32         {d19[1]}, [r8], r12
+        vst1.32         {d22[0]}, [r8], r12
+        vst1.32         {d22[1]}, [r8], r12
+        vst1.32         {d23[0]}, [r8], r12
+        vst1.32         {d23[1]}, [r8], r12
+8:      subs            r3,  r3,  #2
+        add             r0,  r0,  #4
+        it              eq
+        popeq           {r4-r8, pc}
+
+        @ 1 channel
+4:      ldr             r4,  [r1]
+        tst             r2,  #8
+        mov             lr,  r2
+        mov             r5,  r0
+        vld1.32         {q0},     [r4,:128]!
+        vcvt.s32.f32    q0,  q0,  #31
+        vld1.32         {q1},     [r4,:128]!
+        vcvt.s32.f32    q1,  q1,  #31
+        bne             8f
+6:      subs            lr,  lr,  #16
+        vld1.32         {q2},     [r4,:128]!
+        vcvt.s32.f32    q2,  q2,  #31
+        vld1.32         {q3},     [r4,:128]!
+        vcvt.s32.f32    q3,  q3,  #31
+        vst1.16         {d0[1]},  [r5,:16], r12
+        vst1.16         {d0[3]},  [r5,:16], r12
+        vst1.16         {d1[1]},  [r5,:16], r12
+        vst1.16         {d1[3]},  [r5,:16], r12
+        vst1.16         {d2[1]},  [r5,:16], r12
+        vst1.16         {d2[3]},  [r5,:16], r12
+        vst1.16         {d3[1]},  [r5,:16], r12
+        vst1.16         {d3[3]},  [r5,:16], r12
+        beq             7f
+        vld1.32         {q0},     [r4,:128]!
+        vcvt.s32.f32    q0,  q0,  #31
+        vld1.32         {q1},     [r4,:128]!
+        vcvt.s32.f32    q1,  q1,  #31
+7:      vst1.16         {d4[1]},  [r5,:16], r12
+        vst1.16         {d4[3]},  [r5,:16], r12
+        vst1.16         {d5[1]},  [r5,:16], r12
+        vst1.16         {d5[3]},  [r5,:16], r12
+        vst1.16         {d6[1]},  [r5,:16], r12
+        vst1.16         {d6[3]},  [r5,:16], r12
+        vst1.16         {d7[1]},  [r5,:16], r12
+        vst1.16         {d7[3]},  [r5,:16], r12
+        bgt             6b
+        pop             {r4-r8, pc}
+8:      subs            lr,  lr,  #8
+        vst1.16         {d0[1]},  [r5,:16], r12
+        vst1.16         {d0[3]},  [r5,:16], r12
+        vst1.16         {d1[1]},  [r5,:16], r12
+        vst1.16         {d1[3]},  [r5,:16], r12
+        vst1.16         {d2[1]},  [r5,:16], r12
+        vst1.16         {d2[3]},  [r5,:16], r12
+        vst1.16         {d3[1]},  [r5,:16], r12
+        vst1.16         {d3[3]},  [r5,:16], r12
+        it              eq
+        popeq           {r4-r8, pc}
+        vld1.32         {q0},     [r4,:128]!
+        vcvt.s32.f32    q0,  q0,  #31
+        vld1.32         {q1},     [r4,:128]!
+        vcvt.s32.f32    q1,  q1,  #31
+        b               6b
+endfunc
index 1138ef8..e9835c8 100644 (file)
@@ -276,6 +276,8 @@ AudioConvert *ff_audio_convert_alloc(AVAudioResampleContext *avr,
 
     set_generic_function(ac);
 
+    if (ARCH_ARM)
+        ff_audio_convert_init_arm(ac);
     if (ARCH_X86)
         ff_audio_convert_init_x86(ac);
 
index eefffbb..2b8bfac 100644 (file)
@@ -82,6 +82,7 @@ int ff_audio_convert(AudioConvert *ac, AudioData *out, AudioData *in, int len);
 
 /* arch-specific initialization functions */
 
+void ff_audio_convert_init_arm(AudioConvert *ac);
 void ff_audio_convert_init_x86(AudioConvert *ac);
 
 #endif /* AVRESAMPLE_AUDIO_CONVERT_H */
index 61fd395..93bfc0c 100644 (file)
@@ -21,6 +21,8 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <inttypes.h>
+#include <errno.h>
+#include <string.h>
 
 static uint32_t state;
 static uint32_t ran(void)