diff --git a/vp9/common/vp9_rtcd_defs.pl b/vp9/common/vp9_rtcd_defs.pl
index 5df5fc5a8..720e171b5 100644
--- a/vp9/common/vp9_rtcd_defs.pl
+++ b/vp9/common/vp9_rtcd_defs.pl
@@ -141,7 +141,7 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") {
   add_proto qw/void vp9_quantize_fp_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
 
   add_proto qw/void vp9_fdct8x8_quant/, "const int16_t *input, int stride, tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
-  specialize qw/vp9_fdct8x8_quant ssse3/;
+  specialize qw/vp9_fdct8x8_quant neon ssse3/;
 } else {
   add_proto qw/int64_t vp9_block_error/, "const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz";
   specialize qw/vp9_block_error avx2 msa sse2/;
diff --git a/vp9/encoder/arm/neon/vp9_dct_neon.c b/vp9/encoder/arm/neon/vp9_dct_neon.c
index afffc7717..e0b5cc347 100644
--- a/vp9/encoder/arm/neon/vp9_dct_neon.c
+++ b/vp9/encoder/arm/neon/vp9_dct_neon.c
@@ -16,16 +16,17 @@
 
 #include "vp9/common/vp9_blockd.h"
 #include "vpx_dsp/txfm_common.h"
+#include "vpx_dsp/vpx_dsp_common.h"
 
 void vp9_fdct8x8_quant_neon(const int16_t *input, int stride,
-                            int16_t *coeff_ptr, intptr_t n_coeffs,
+                            tran_low_t *coeff_ptr, intptr_t n_coeffs,
                             int skip_block, const int16_t *zbin_ptr,
                             const int16_t *round_ptr, const int16_t *quant_ptr,
-                            const int16_t *quant_shift_ptr, int16_t *qcoeff_ptr,
-                            int16_t *dqcoeff_ptr, const int16_t *dequant_ptr,
-                            uint16_t *eob_ptr, const int16_t *scan_ptr,
-                            const int16_t *iscan_ptr) {
-  int16_t temp_buffer[64];
+                            const int16_t *quant_shift_ptr,
+                            tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr,
+                            const int16_t *dequant_ptr, uint16_t *eob_ptr,
+                            const int16_t *scan_ptr, const int16_t *iscan_ptr) {
+  tran_low_t temp_buffer[64];
   (void)coeff_ptr;
 
   vpx_fdct8x8_neon(input, temp_buffer, stride);
diff --git a/vp9/vp9cx.mk b/vp9/vp9cx.mk
index ad33fa1b4..a54e99e2c 100644
--- a/vp9/vp9cx.mk
+++ b/vp9/vp9cx.mk
@@ -130,9 +130,9 @@ endif
 VP9_CX_SRCS-$(HAVE_AVX2) += encoder/x86/vp9_error_intrin_avx2.c
 
 ifneq ($(CONFIG_VP9_HIGHBITDEPTH),yes)
-VP9_CX_SRCS-$(HAVE_NEON) += encoder/arm/neon/vp9_dct_neon.c
 VP9_CX_SRCS-$(HAVE_NEON) += encoder/arm/neon/vp9_error_neon.c
 endif
+VP9_CX_SRCS-$(HAVE_NEON) += encoder/arm/neon/vp9_dct_neon.c
 VP9_CX_SRCS-$(HAVE_NEON) += encoder/arm/neon/vp9_quantize_neon.c
 
 VP9_CX_SRCS-$(HAVE_MSA) += encoder/mips/msa/vp9_error_msa.c