*_neon.c: make some functions static

fixes some -Wmissing-prototypes warnings

Bug: aomedia:3416
Change-Id: I571e0f5329808a3129a3b66f54d579da72e0eb86

aom_dsp/arm/blend_a64_mask_neon.c

diff --git a/aom_dsp/arm/blend_a64_mask_neon.c b/aom_dsp/arm/blend_a64_mask_neon.c
index 1bc3b80..48ff683 100644
--- a/aom_dsp/arm/blend_a64_mask_neon.c
+++ b/aom_dsp/arm/blend_a64_mask_neon.c
@@ -20,8 +20,9 @@
 #include "aom_dsp/arm/mem_neon.h"
 #include "aom_dsp/blend.h"
 
-uint8x8_t alpha_blend_a64_d16_u16x8(uint16x8_t m, uint16x8_t a, uint16x8_t b,
-                                    uint16x8_t round_offset) {
+static uint8x8_t alpha_blend_a64_d16_u16x8(uint16x8_t m, uint16x8_t a,
+                                           uint16x8_t b,
+                                           uint16x8_t round_offset) {
   const uint16x8_t m_inv = vsubq_u16(vdupq_n_u16(AOM_BLEND_A64_MAX_ALPHA), m);
 
   uint32x4_t blend_u32_lo = vmull_u16(vget_low_u16(m), vget_low_u16(a));

aom_dsp/arm/highbd_intrapred_neon.c

diff --git a/aom_dsp/arm/highbd_intrapred_neon.c b/aom_dsp/arm/highbd_intrapred_neon.c
index dc47974..d5a0044 100644
--- a/aom_dsp/arm/highbd_intrapred_neon.c
+++ b/aom_dsp/arm/highbd_intrapred_neon.c
@@ -1201,7 +1201,7 @@
 
 // For width 16 and above.
 #define HIGHBD_SMOOTH_H_PREDICTOR(W)                                          \
-  void highbd_smooth_h_##W##xh_neon(                                          \
+  static void highbd_smooth_h_##W##xh_neon(                                   \
       uint16_t *dst, ptrdiff_t stride, const uint16_t *const top_row,         \
       const uint16_t *const left_column, const int height) {                  \
     const uint16_t top_right = top_row[(W)-1];                                \

av1/common/arm/av1_inv_txfm_neon.c

diff --git a/av1/common/arm/av1_inv_txfm_neon.c b/av1/common/arm/av1_inv_txfm_neon.c
index 09e5166..4a66b90 100644
--- a/av1/common/arm/av1_inv_txfm_neon.c
+++ b/av1/common/arm/av1_inv_txfm_neon.c
@@ -447,7 +447,7 @@
   out[7] = step1;
 }
 
-void av1_round_shift_array_16_neon(int16x8_t *arr, int size, int bit) {
+static void round_shift_array_16_neon(int16x8_t *arr, int size, int bit) {
   assert(!(size % 4));
   if (!bit) return;
   const int16x8_t dup_bits_n_16x8 = vdupq_n_s16((int16_t)(-bit));
@@ -3661,7 +3661,7 @@
       round_shift_for_rect(cur_a, cur_a, buf_size_nonzero_w);
     }
     row_txfm(cur_a, cur_a, INV_COS_BIT);
-    av1_round_shift_array_16_neon(cur_a, txfm_size_col, -shift[0]);
+    round_shift_array_16_neon(cur_a, txfm_size_col, -shift[0]);
     if (lr_flip == 1) {
       for (int j = 0; j < buf_size_w_div8; ++j) {
         flip_buf_ud_neon(&cur_a[j * 8], 8);
@@ -3736,8 +3736,7 @@
   }
   for (int j = 0; j < buf_size_w_div8; ++j) {
     col_txfm(&b[j * txfm_size_row], &b[j * txfm_size_row], INV_COS_BIT);
-    av1_round_shift_array_16_neon(&b[j * txfm_size_row], txfm_size_row,
-                                  -shift[1]);
+    round_shift_array_16_neon(&b[j * txfm_size_row], txfm_size_row, -shift[1]);
   }
   if (txfm_size_col >= 16) {
     for (int i = 0; i < (txfm_size_col >> 4); i++) {
@@ -4112,7 +4111,7 @@
       round_shift_for_rect(cur_a, cur_a, buf_size_nonzero_w);
     }
     row_txfm(cur_a, cur_a, INV_COS_BIT);
-    av1_round_shift_array_16_neon(cur_a, txfm_size_col, -shift[0]);
+    round_shift_array_16_neon(cur_a, txfm_size_col, -shift[0]);
     if (lr_flip == 1) {
       for (int j = 0; j < buf_size_w_div8; ++j) {
         flip_buf_ud_neon(&cur_a[j * 8], 8);
@@ -4130,8 +4129,7 @@
   }
   for (int j = 0; j < buf_size_w_div8; ++j) {
     col_txfm(&b[j * txfm_size_row], &b[j * txfm_size_row], INV_COS_BIT);
-    av1_round_shift_array_16_neon(&b[j * txfm_size_row], txfm_size_row,
-                                  -shift[1]);
+    round_shift_array_16_neon(&b[j * txfm_size_row], txfm_size_row, -shift[1]);
   }
 
   if (txfm_size_col >= 16) {

av1/common/arm/selfguided_neon.c

diff --git a/av1/common/arm/selfguided_neon.c b/av1/common/arm/selfguided_neon.c
index 1d3a3cc..08e298f 100644
--- a/av1/common/arm/selfguided_neon.c
+++ b/av1/common/arm/selfguided_neon.c
@@ -1124,10 +1124,10 @@
   } while (h > 0);
 }
 
-void final_filter_internal(uint16_t *A, int32_t *B, const int buf_stride,
-                           int16_t *src, const int src_stride, int32_t *dst,
-                           const int dst_stride, const int width,
-                           const int height) {
+static void final_filter_internal(uint16_t *A, int32_t *B, const int buf_stride,
+                                  int16_t *src, const int src_stride,
+                                  int32_t *dst, const int dst_stride,
+                                  const int width, const int height) {
   int16x8_t s0;
   int32_t *B_tmp, *dst_ptr;
   uint16_t *A_tmp;