[FFmpeg-devel] avutil/mips: refactor msa load and store macros.

Submitted by Shiyou Yin on July 17, 2019, 9:35 a.m.

Details

Message ID 1563356100-3518-1-git-send-email-yinshiyou-hf@loongson.cn
State Accepted
Commit 153c60752558369b98dce0b7a0ca7acc687fa630
Headers show

Commit Message

Shiyou Yin July 17, 2019, 9:35 a.m.
Replace STnxm_UB and LDnxm_SH with new macros ST_{H/W/D}{1/2/4/8}.
The old macros are difficult to use because they don't follow the same parameter passing rules.
Changing details as following:
1. remove LD4x4_SH.
2. replace ST2x4_UB with ST_H4.
3. replace ST4x2_UB with ST_W2.
4. replace ST4x4_UB with ST_W4.
5. replace ST4x8_UB with ST_W8.
6. replace ST6x4_UB with ST_W2 and ST_H2.
7. replace ST8x1_UB with ST_D1.
8. replace ST8x2_UB with ST_D2.
9. replace ST8x4_UB with ST_D4.
10. replace ST8x8_UB with ST_D8.
11. replace ST12x4_UB with ST_D4 and ST_W4.

Examples of new macro: ST_H4(in, idx0, idx1, idx2, idx3, pdst, stride)
ST_H4 store four half-word elements in vector 'in' to pdst with stride.
About the macro name:
1) 'ST' means store operation.
2) 'H/W/D' means type of vector element is 'half-word/word/double-word'.
3) Number '1/2/4/8' means how many elements will be stored.
About the macro parameter:
1) 'in0, in1...' 128-bits vector.
2) 'idx0, idx1...' elements index.
3) 'pdst' destination pointer to store to
4) 'stride' stride of each store operation.
---
 libavcodec/mips/h263dsp_msa.c       |   5 +-
 libavcodec/mips/h264chroma_msa.c    |  87 +++++-----
 libavcodec/mips/h264dsp_msa.c       |  50 +++---
 libavcodec/mips/h264idct_msa.c      |  10 +-
 libavcodec/mips/h264qpel_msa.c      | 138 ++++++++--------
 libavcodec/mips/hevc_idct_msa.c     |   7 +-
 libavcodec/mips/hevc_lpf_sao_msa.c  |  32 ++--
 libavcodec/mips/hevc_mc_bi_msa.c    | 135 +++++++++-------
 libavcodec/mips/hevc_mc_biw_msa.c   | 128 ++++++++-------
 libavcodec/mips/hevc_mc_uni_msa.c   | 146 ++++++++---------
 libavcodec/mips/hevc_mc_uniw_msa.c  | 157 ++++++++++--------
 libavcodec/mips/hevcdsp_msa.c       |  63 ++++----
 libavcodec/mips/hevcpred_msa.c      |  45 +++---
 libavcodec/mips/hpeldsp_msa.c       |  10 +-
 libavcodec/mips/qpeldsp_msa.c       | 292 ++++++++++++++-------------------
 libavcodec/mips/vp3dsp_idct_msa.c   |  60 +++----
 libavcodec/mips/vp8_idct_msa.c      |   6 +-
 libavcodec/mips/vp8_lpf_msa.c       |  27 +---
 libavcodec/mips/vp8_mc_msa.c        | 109 +++++++------
 libavcodec/mips/vp9_idct_msa.c      |  70 ++++----
 libavcodec/mips/vp9_intra_msa.c     |   5 +-
 libavcodec/mips/vp9_lpf_msa.c       | 143 ++++++++---------
 libavcodec/mips/vp9_mc_msa.c        | 105 ++++++------
 libavutil/mips/generic_macros_msa.h | 312 +++++++++++++-----------------------
 24 files changed, 991 insertions(+), 1151 deletions(-)

Comments

Michael Niedermayer July 19, 2019, 8:21 a.m.
On Wed, Jul 17, 2019 at 05:35:00PM +0800, Shiyou Yin wrote:
> Replace STnxm_UB and LDnxm_SH with new macros ST_{H/W/D}{1/2/4/8}.
> The old macros are difficult to use because they don't follow the same parameter passing rules.
> Changing details as following:
> 1. remove LD4x4_SH.
> 2. replace ST2x4_UB with ST_H4.
> 3. replace ST4x2_UB with ST_W2.
> 4. replace ST4x4_UB with ST_W4.
> 5. replace ST4x8_UB with ST_W8.
> 6. replace ST6x4_UB with ST_W2 and ST_H2.
> 7. replace ST8x1_UB with ST_D1.
> 8. replace ST8x2_UB with ST_D2.
> 9. replace ST8x4_UB with ST_D4.
> 10. replace ST8x8_UB with ST_D8.
> 11. replace ST12x4_UB with ST_D4 and ST_W4.
> 
> Examples of new macro: ST_H4(in, idx0, idx1, idx2, idx3, pdst, stride)
> ST_H4 store four half-word elements in vector 'in' to pdst with stride.
> About the macro name:
> 1) 'ST' means store operation.
> 2) 'H/W/D' means type of vector element is 'half-word/word/double-word'.
> 3) Number '1/2/4/8' means how many elements will be stored.
> About the macro parameter:
> 1) 'in0, in1...' 128-bits vector.
> 2) 'idx0, idx1...' elements index.
> 3) 'pdst' destination pointer to store to
> 4) 'stride' stride of each store operation.

will apply

thanks

[...]

Patch hide | download patch | download mbox

diff --git a/libavcodec/mips/h263dsp_msa.c b/libavcodec/mips/h263dsp_msa.c
index 472bcbd..2e1ca01 100644
--- a/libavcodec/mips/h263dsp_msa.c
+++ b/libavcodec/mips/h263dsp_msa.c
@@ -86,10 +86,7 @@  static void h263_h_loop_filter_msa(uint8_t *src, int32_t stride, int32_t qscale)
     ILVR_B2_SH(in3, in0, in1, in2, temp0, temp1);
     in0 = (v16u8) __msa_ilvr_h(temp1, temp0);
     in3 = (v16u8) __msa_ilvl_h(temp1, temp0);
-    ST4x4_UB(in0, in0, 0, 1, 2, 3, src, stride);
-    src += 4 * stride;
-    ST4x4_UB(in3, in3, 0, 1, 2, 3, src, stride);
-    src += 4 * stride;
+    ST_W8(in0, in3, 0, 1, 2, 3, 0, 1, 2, 3, src, stride);
 }
 
 static void h263_v_loop_filter_msa(uint8_t *src, int32_t stride, int32_t qscale)
diff --git a/libavcodec/mips/h264chroma_msa.c b/libavcodec/mips/h264chroma_msa.c
index 4c25761..4a68d9e 100644
--- a/libavcodec/mips/h264chroma_msa.c
+++ b/libavcodec/mips/h264chroma_msa.c
@@ -85,7 +85,7 @@  static void avc_chroma_hz_2x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     res_r = __msa_sat_u_h(res_r, 7);
     res = (v8i16) __msa_pckev_b((v16i8) res_r, (v16i8) res_r);
 
-    ST2x4_UB(res, 0, dst, stride);
+    ST_H4(res, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_hz_2w_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -121,7 +121,7 @@  static void avc_chroma_hz_4x2_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     res_r = __msa_sat_u_h(res_r, 7);
     res = (v4i32) __msa_pckev_b((v16i8) res_r, (v16i8) res_r);
 
-    ST4x2_UB(res, dst, stride);
+    ST_W2(res, 0, 1, dst, stride);
 }
 
 static void avc_chroma_hz_4x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -144,7 +144,7 @@  static void avc_chroma_hz_4x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SRARI_H2_UH(res0_r, res1_r, 6);
     SAT_UH2_UH(res0_r, res1_r, 7);
     out = (v16u8) __msa_pckev_b((v16i8) res1_r, (v16i8) res0_r);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_hz_4x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -168,7 +168,7 @@  static void avc_chroma_hz_4x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SRARI_H4_UH(res0, res1, res2, res3, 6);
     SAT_UH4_UH(res0, res1, res2, res3, 7);
     PCKEV_B2_UB(res1, res0, res3, res2, out0, out1);
-    ST4x8_UB(out0, out1, dst, stride);
+    ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_hz_4w_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -204,7 +204,7 @@  static void avc_chroma_hz_8x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SRARI_H4_UH(res0, res1, res2, res3, 6);
     SAT_UH4_UH(res0, res1, res2, res3, 7);
     PCKEV_B2_UB(res1, res0, res3, res2, out0, out1);
-    ST8x4_UB(out0, out1, dst, stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_chroma_hz_8x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -237,7 +237,7 @@  static void avc_chroma_hz_8x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SAT_UH4_UH(res4, res5, res6, res7, 7);
     PCKEV_B2_UB(res1, res0, res3, res2, out0, out1);
     PCKEV_B2_UB(res5, res4, res7, res6, out2, out3);
-    ST8x8_UB(out0, out1, out2, out3, dst, stride);
+    ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_chroma_hz_nonmult_msa(uint8_t *src, uint8_t *dst,
@@ -266,7 +266,7 @@  static void avc_chroma_hz_nonmult_msa(uint8_t *src, uint8_t *dst,
         SRARI_H4_UH(res0, res1, res2, res3, 6);
         SAT_UH4_UH(res0, res1, res2, res3, 7);
         PCKEV_B2_UB(res1, res0, res3, res2, out0, out1);
-        ST8x4_UB(out0, out1, dst, stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
         dst += (4 * stride);
     }
 
@@ -283,7 +283,7 @@  static void avc_chroma_hz_nonmult_msa(uint8_t *src, uint8_t *dst,
             res0 = __msa_sat_u_h(res0, 7);
             res0 = (v8u16) __msa_pckev_b((v16i8) res0, (v16i8) res0);
 
-            ST8x1_UB(res0, dst);
+            ST_D1(res0, 0, dst);
             dst += stride;
         }
     }
@@ -359,7 +359,7 @@  static void avc_chroma_vt_2x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
 
     res = (v8i16) __msa_pckev_b((v16i8) res_r, (v16i8) res_r);
 
-    ST2x4_UB(res, 0, dst, stride);
+    ST_H4(res, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_vt_2w_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -394,7 +394,7 @@  static void avc_chroma_vt_4x2_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     res_r = __msa_sat_u_h(res_r, 7);
     res = (v4i32) __msa_pckev_b((v16i8) res_r, (v16i8) res_r);
 
-    ST4x2_UB(res, dst, stride);
+    ST_W2(res, 0, 1, dst, stride);
 }
 
 static void avc_chroma_vt_4x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -418,7 +418,7 @@  static void avc_chroma_vt_4x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SRARI_H2_UH(res0_r, res1_r, 6);
     SAT_UH2_UH(res0_r, res1_r, 7);
     out = (v16u8) __msa_pckev_b((v16i8) res1_r, (v16i8) res0_r);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_vt_4x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -446,7 +446,7 @@  static void avc_chroma_vt_4x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SRARI_H4_UH(res0, res1, res2, res3, 6);
     SAT_UH4_UH(res0, res1, res2, res3, 7);
     PCKEV_B2_UB(res1, res0, res3, res2, out0, out1);
-    ST4x8_UB(out0, out1, dst, stride);
+    ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_vt_4w_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -480,7 +480,7 @@  static void avc_chroma_vt_8x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SRARI_H4_UH(res0, res1, res2, res3, 6);
     SAT_UH4_UH(res0, res1, res2, res3, 7);
     PCKEV_B2_UB(res1, res0, res3, res2, out0, out1);
-    ST8x4_UB(out0, out1, dst, stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_chroma_vt_8x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -512,7 +512,7 @@  static void avc_chroma_vt_8x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SAT_UH4_UH(res0, res1, res2, res3, 7);
     PCKEV_B2_UB(res1, res0, res3, res2, out0, out1);
     PCKEV_B2_UB(res5, res4, res7, res6, out2, out3);
-    ST8x8_UB(out0, out1, out2, out3, dst, stride);
+    ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_chroma_vt_8w_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -592,7 +592,7 @@  static void avc_chroma_hv_2x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
 
     res = (v8i16) __msa_pckev_b((v16i8) res_vt0, (v16i8) res_vt0);
 
-    ST2x4_UB(res, 0, dst, stride);
+    ST_H4(res, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_hv_2w_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -634,7 +634,7 @@  static void avc_chroma_hv_4x2_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     res_vt0 = __msa_sat_u_h(res_vt0, 7);
     res = (v4i32) __msa_pckev_b((v16i8) res_vt0, (v16i8) res_vt0);
 
-    ST4x2_UB(res, dst, stride);
+    ST_W2(res, 0, 1, dst, stride);
 }
 
 static void avc_chroma_hv_4x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -666,7 +666,8 @@  static void avc_chroma_hv_4x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SRARI_H2_UH(res_vt0, res_vt1, 6);
     SAT_UH2_UH(res_vt0, res_vt1, 7);
     PCKEV_B2_SW(res_vt0, res_vt0, res_vt1, res_vt1, res0, res1);
-    ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, stride);
+    ST_W2(res0, 0, 1, dst, stride);
+    ST_W2(res1, 0, 1, dst + 2 * stride, stride);
 }
 
 static void avc_chroma_hv_4x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -706,7 +707,7 @@  static void avc_chroma_hv_4x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SRARI_H4_UH(res_vt0, res_vt1, res_vt2, res_vt3, 6);
     SAT_UH4_UH(res_vt0, res_vt1, res_vt2, res_vt3, 7);
     PCKEV_B2_UB(res_vt1, res_vt0, res_vt3, res_vt2, res0, res1);
-    ST4x8_UB(res0, res1, dst, stride);
+    ST_W8(res0, res1, 0, 1, 2, 3, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_hv_4w_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -766,7 +767,7 @@  static void avc_chroma_hv_8x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SRARI_H4_UH(res_vt0, res_vt1, res_vt2, res_vt3, 6);
     SAT_UH4_UH(res_vt0, res_vt1, res_vt2, res_vt3, 7);
     PCKEV_B2_UB(res_vt1, res_vt0, res_vt3, res_vt2, out0, out1);
-    ST8x4_UB(out0, out1, dst, stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_chroma_hv_8x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -822,7 +823,7 @@  static void avc_chroma_hv_8x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SAT_UH4_UH(res_vt4, res_vt5, res_vt6, res_vt7, 7);
     PCKEV_B2_UB(res_vt1, res_vt0, res_vt3, res_vt2, out0, out1);
     PCKEV_B2_UB(res_vt5, res_vt4, res_vt7, res_vt6, out2, out3);
-    ST8x8_UB(out0, out1, out2, out3, dst, stride);
+    ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_chroma_hv_8w_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -918,7 +919,7 @@  static void avc_chroma_hz_and_aver_dst_2x4_msa(uint8_t *src, uint8_t *dst,
     dst0 = (v16u8) __msa_pckev_b((v16i8) res_r, (v16i8) res_r);
     dst0 = __msa_aver_u_b(dst0, dst_data);
 
-    ST2x4_UB(dst0, 0, dst, stride);
+    ST_H4(dst0, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_hz_and_aver_dst_2w_msa(uint8_t *src, uint8_t *dst,
@@ -962,7 +963,7 @@  static void avc_chroma_hz_and_aver_dst_4x2_msa(uint8_t *src, uint8_t *dst,
     res = __msa_pckev_b((v16i8) res_r, (v16i8) res_r);
     dst_data = __msa_aver_u_b((v16u8) res, dst_data);
 
-    ST4x2_UB(dst_data, dst, stride);
+    ST_W2(dst_data, 0, 1, dst, stride);
 }
 
 static void avc_chroma_hz_and_aver_dst_4x4_msa(uint8_t *src, uint8_t *dst,
@@ -991,7 +992,7 @@  static void avc_chroma_hz_and_aver_dst_4x4_msa(uint8_t *src, uint8_t *dst,
     SAT_UH2_UH(res0_r, res1_r, 7);
     out = (v16u8) __msa_pckev_b((v16i8) res1_r, (v16i8) res0_r);
     out = __msa_aver_u_b(out, dst_data);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_hz_and_aver_dst_4x8_msa(uint8_t *src, uint8_t *dst,
@@ -1023,7 +1024,7 @@  static void avc_chroma_hz_and_aver_dst_4x8_msa(uint8_t *src, uint8_t *dst,
     SAT_UH4_UH(res0, res1, res2, res3, 7);
     PCKEV_B2_UB(res1, res0, res3, res2, out0, out1);
     AVER_UB2_UB(out0, dst0, out1, dst1, out0, out1);
-    ST4x8_UB(out0, out1, dst, stride);
+    ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_hz_and_aver_dst_4w_msa(uint8_t *src, uint8_t *dst,
@@ -1066,7 +1067,7 @@  static void avc_chroma_hz_and_aver_dst_8x4_msa(uint8_t *src, uint8_t *dst,
     SAT_UH4_UH(res0, res1, res2, res3, 7);
     PCKEV_B2_UB(res1, res0, res3, res2, out0, out1);
     AVER_UB2_UB(out0, dst0, out1, dst1, dst0, dst1);
-    ST8x4_UB(dst0, dst1, dst, stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_chroma_hz_and_aver_dst_8x8_msa(uint8_t *src, uint8_t *dst,
@@ -1110,7 +1111,7 @@  static void avc_chroma_hz_and_aver_dst_8x8_msa(uint8_t *src, uint8_t *dst,
     PCKEV_B2_UB(res5, res4, res7, res6, out2, out3);
     AVER_UB2_UB(out0, dst0, out1, dst1, out0, out1);
     AVER_UB2_UB(out2, dst2, out3, dst3, out2, out3);
-    ST8x8_UB(out0, out1, out2, out3, dst, stride);
+    ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_chroma_hz_and_aver_dst_8w_msa(uint8_t *src, uint8_t *dst,
@@ -1200,7 +1201,7 @@  static void avc_chroma_vt_and_aver_dst_2x4_msa(uint8_t *src, uint8_t *dst,
     res = (v8i16) __msa_pckev_b((v16i8) res_r, (v16i8) res_r);
     res = (v8i16) __msa_aver_u_b((v16u8) res, dst_data);
 
-    ST2x4_UB(res, 0, dst, stride);
+    ST_H4(res, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_vt_and_aver_dst_2w_msa(uint8_t *src, uint8_t *dst,
@@ -1243,7 +1244,7 @@  static void avc_chroma_vt_and_aver_dst_4x2_msa(uint8_t *src, uint8_t *dst,
     res = (v16u8) __msa_pckev_b((v16i8) res_r, (v16i8) res_r);
     res = __msa_aver_u_b(res, dst_data);
 
-    ST4x2_UB(res, dst, stride);
+    ST_W2(res, 0, 1, dst, stride);
 }
 
 static void avc_chroma_vt_and_aver_dst_4x4_msa(uint8_t *src, uint8_t *dst,
@@ -1273,7 +1274,7 @@  static void avc_chroma_vt_and_aver_dst_4x4_msa(uint8_t *src, uint8_t *dst,
     SAT_UH2_UH(res0_r, res1_r, 7);
     out = (v16u8) __msa_pckev_b((v16i8) res1_r, (v16i8) res0_r);
     out = __msa_aver_u_b(out, dst0);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_vt_and_aver_dst_4x8_msa(uint8_t *src, uint8_t *dst,
@@ -1309,7 +1310,7 @@  static void avc_chroma_vt_and_aver_dst_4x8_msa(uint8_t *src, uint8_t *dst,
     SAT_UH4_UH(res0, res1, res2, res3, 7);
     PCKEV_B2_UB(res1, res0, res3, res2, out0, out1);
     AVER_UB2_UB(out0, dst0, out1, dst1, out0, out1);
-    ST4x8_UB(out0, out1, dst, stride);
+    ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_vt_and_aver_dst_4w_msa(uint8_t *src, uint8_t *dst,
@@ -1351,7 +1352,7 @@  static void avc_chroma_vt_and_aver_dst_8x4_msa(uint8_t *src, uint8_t *dst,
     SAT_UH4_UH(res0, res1, res2, res3, 7);
     PCKEV_B2_UB(res1, res0, res3, res2, out0, out1);
     AVER_UB2_UB(out0, dst0, out1, dst1, out0, out1);
-    ST8x4_UB(out0, out1, dst, stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_chroma_vt_and_aver_dst_8x8_msa(uint8_t *src, uint8_t *dst,
@@ -1394,7 +1395,7 @@  static void avc_chroma_vt_and_aver_dst_8x8_msa(uint8_t *src, uint8_t *dst,
     PCKEV_B2_UB(res5, res4, res7, res6, out2, out3);
     AVER_UB2_UB(out0, dst0, out1, dst1, out0, out1);
     AVER_UB2_UB(out2, dst2, out3, dst3, out2, out3);
-    ST8x8_UB(out0, out1, out2, out3, dst, stride);
+    ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_chroma_vt_and_aver_dst_8w_msa(uint8_t *src, uint8_t *dst,
@@ -1492,7 +1493,7 @@  static void avc_chroma_hv_and_aver_dst_2x4_msa(uint8_t *src, uint8_t *dst,
     res = __msa_pckev_b((v16i8) res_vt0, (v16i8) res_vt0);
     dst0 = __msa_aver_u_b((v16u8) res, dst0);
 
-    ST2x4_UB(dst0, 0, dst, stride);
+    ST_H4(dst0, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_hv_and_aver_dst_2w_msa(uint8_t *src, uint8_t *dst,
@@ -1545,7 +1546,7 @@  static void avc_chroma_hv_and_aver_dst_4x2_msa(uint8_t *src, uint8_t *dst,
     dst0 = (v16u8) __msa_pckev_b((v16i8) res_vt0, (v16i8) res_vt0);
     dst0 = __msa_aver_u_b(dst0, dst_data);
 
-    ST4x2_UB(dst0, dst, stride);
+    ST_W2(dst0, 0, 1, dst, stride);
 }
 
 static void avc_chroma_hv_and_aver_dst_4x4_msa(uint8_t *src, uint8_t *dst,
@@ -1584,7 +1585,7 @@  static void avc_chroma_hv_and_aver_dst_4x4_msa(uint8_t *src, uint8_t *dst,
     SAT_UH2_UH(res_vt0, res_vt1, 7);
     out = (v16u8) __msa_pckev_b((v16i8) res_vt1, (v16i8) res_vt0);
     out = __msa_aver_u_b(out, dst_data);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_hv_and_aver_dst_4x8_msa(uint8_t *src, uint8_t *dst,
@@ -1633,7 +1634,7 @@  static void avc_chroma_hv_and_aver_dst_4x8_msa(uint8_t *src, uint8_t *dst,
     SAT_UH4_UH(res_vt0, res_vt1, res_vt2, res_vt3, 7);
     PCKEV_B2_UB(res_vt1, res_vt0, res_vt3, res_vt2, res0, res1);
     AVER_UB2_UB(res0, dst0, res1, dst1, res0, res1);
-    ST4x8_UB(res0, res1, dst, stride);
+    ST_W8(res0, res1, 0, 1, 2, 3, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_chroma_hv_and_aver_dst_4w_msa(uint8_t *src, uint8_t *dst,
@@ -1701,7 +1702,7 @@  static void avc_chroma_hv_and_aver_dst_8x4_msa(uint8_t *src, uint8_t *dst,
     SAT_UH4_UH(res_vt0, res_vt1, res_vt2, res_vt3, 7);
     PCKEV_B2_UB(res_vt1, res_vt0, res_vt3, res_vt2, out0, out1);
     AVER_UB2_UB(out0, dst0, out1, dst1, out0, out1);
-    ST8x4_UB(out0, out1, dst, stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_chroma_hv_and_aver_dst_8x8_msa(uint8_t *src, uint8_t *dst,
@@ -1770,7 +1771,7 @@  static void avc_chroma_hv_and_aver_dst_8x8_msa(uint8_t *src, uint8_t *dst,
     PCKEV_B2_UB(res_vt5, res_vt4, res_vt7, res_vt6, out2, out3);
     AVER_UB2_UB(out0, dst0, out1, dst1, out0, out1);
     AVER_UB2_UB(out2, dst2, out3, dst3, out2, out3);
-    ST8x8_UB(out0, out1, out2, out3, dst, stride);
+    ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_chroma_hv_and_aver_dst_8w_msa(uint8_t *src, uint8_t *dst,
@@ -1848,21 +1849,21 @@  static void avg_width4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
         LW4(dst + 4 * stride, stride, tp0, tp1, tp2, tp3);
         INSERT_W4_UB(tp0, tp1, tp2, tp3, dst1);
         AVER_UB2_UB(src0, dst0, src1, dst1, dst0, dst1);
-        ST4x8_UB(dst0, dst1, dst, stride);
+        ST_W8(dst0, dst1, 0, 1, 2, 3, 0, 1, 2, 3, dst, stride);
     } else if (4 == height) {
         LW4(src, stride, tp0, tp1, tp2, tp3);
         INSERT_W4_UB(tp0, tp1, tp2, tp3, src0);
         LW4(dst, stride, tp0, tp1, tp2, tp3);
         INSERT_W4_UB(tp0, tp1, tp2, tp3, dst0);
         dst0 = __msa_aver_u_b(src0, dst0);
-        ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, stride);
+        ST_W4(dst0, 0, 1, 2, 3, dst, stride);
     } else if (2 == height) {
         LW2(src, stride, tp0, tp1);
         INSERT_W2_UB(tp0, tp1, src0);
         LW2(dst, stride, tp0, tp1);
         INSERT_W2_UB(tp0, tp1, dst0);
         dst0 = __msa_aver_u_b(src0, dst0);
-        ST4x2_UB(dst0, dst, stride);
+        ST_W2(dst0, 0, 1, dst, stride);
     }
 }
 
@@ -1889,7 +1890,7 @@  static void avg_width8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
         INSERT_D2_UB(tp6, tp7, dst3);
         AVER_UB4_UB(src0, dst0, src1, dst1, src2, dst2, src3, dst3, dst0, dst1,
                     dst2, dst3);
-        ST8x8_UB(dst0, dst1, dst2, dst3, dst, stride);
+        ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
     } else if (4 == height) {
         LD4(src, stride, tp0, tp1, tp2, tp3);
         INSERT_D2_UB(tp0, tp1, src0);
@@ -1898,7 +1899,7 @@  static void avg_width8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
         INSERT_D2_UB(tp0, tp1, dst0);
         INSERT_D2_UB(tp2, tp3, dst1);
         AVER_UB2_UB(src0, dst0, src1, dst1, dst0, dst1);
-        ST8x4_UB(dst0, dst1, dst, stride);
+        ST_D4(dst0, dst1, 0, 1, 0, 1, dst, stride);
     }
 }
 
diff --git a/libavcodec/mips/h264dsp_msa.c b/libavcodec/mips/h264dsp_msa.c
index e50f5ca..89fe399 100644
--- a/libavcodec/mips/h264dsp_msa.c
+++ b/libavcodec/mips/h264dsp_msa.c
@@ -45,7 +45,7 @@  static void avc_wgt_4x2_msa(uint8_t *data, int32_t stride,
     tmp0 = __msa_srlr_h(tmp0, denom);
     tmp0 = (v8i16) __msa_sat_u_h((v8u16) tmp0, 7);
     src0 = (v16u8) __msa_pckev_b((v16i8) tmp0, (v16i8) tmp0);
-    ST4x2_UB(src0, data, stride);
+    ST_W2(src0, 0, 1, data, stride);
 }
 
 static void avc_wgt_4x4_msa(uint8_t *data, int32_t stride, int32_t log2_denom,
@@ -71,7 +71,7 @@  static void avc_wgt_4x4_msa(uint8_t *data, int32_t stride, int32_t log2_denom,
     tmp1 = __msa_srlr_h(tmp1, denom);
     SAT_UH2_SH(tmp0, tmp1, 7);
     src0 = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-    ST4x4_UB(src0, src0, 0, 1, 2, 3, data, stride);
+    ST_W4(src0, 0, 1, 2, 3, data, stride);
 }
 
 static void avc_wgt_4x8_msa(uint8_t *data, int32_t stride, int32_t log2_denom,
@@ -102,7 +102,7 @@  static void avc_wgt_4x8_msa(uint8_t *data, int32_t stride, int32_t log2_denom,
     SRLR_H4_SH(tmp0, tmp1, tmp2, tmp3, denom);
     SAT_UH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
     PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, src0, src1);
-    ST4x8_UB(src0, src1, data, stride);
+    ST_W8(src0, src1, 0, 1, 2, 3, 0, 1, 2, 3, data, stride);
 }
 
 static void avc_wgt_8x4_msa(uint8_t *data, int32_t stride, int32_t log2_denom,
@@ -133,7 +133,7 @@  static void avc_wgt_8x4_msa(uint8_t *data, int32_t stride, int32_t log2_denom,
     SRLR_H4_SH(tmp0, tmp1, tmp2, tmp3, denom);
     SAT_UH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
     PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, src0, src1);
-    ST8x4_UB(src0, src1, data, stride);
+    ST_D4(src0, src1, 0, 1, 0, 1, data, stride);
 }
 
 static void avc_wgt_8x8_msa(uint8_t *data, int32_t stride, int32_t log2_denom,
@@ -175,7 +175,7 @@  static void avc_wgt_8x8_msa(uint8_t *data, int32_t stride, int32_t log2_denom,
     SAT_UH8_SH(tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7, 7);
     PCKEV_B4_UB(tmp1, tmp0, tmp3, tmp2, tmp5, tmp4, tmp7, tmp6, src0, src1,
                 src2, src3);
-    ST8x8_UB(src0, src1, src2, src3, data, stride);
+    ST_D8(src0, src1, src2, src3, 0, 1, 0, 1, 0, 1, 0, 1, data, stride);
 }
 
 static void avc_wgt_8x16_msa(uint8_t *data, int32_t stride, int32_t log2_denom,
@@ -218,7 +218,7 @@  static void avc_wgt_8x16_msa(uint8_t *data, int32_t stride, int32_t log2_denom,
         SAT_UH8_SH(tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7, 7);
         PCKEV_B4_UB(tmp1, tmp0, tmp3, tmp2, tmp5, tmp4, tmp7, tmp6, src0, src1,
                     src2, src3);
-        ST8x8_UB(src0, src1, src2, src3, data, stride);
+        ST_D8(src0, src1, src2, src3, 0, 1, 0, 1, 0, 1, 0, 1, data, stride);
         data += 8 * stride;
     }
 }
@@ -253,7 +253,7 @@  static void avc_biwgt_4x2_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     tmp0 = __msa_maxi_s_h(tmp0, 0);
     tmp0 = __msa_min_s_h(max255, tmp0);
     dst0 = (v16u8) __msa_pckev_b((v16i8) tmp0, (v16i8) tmp0);
-    ST4x2_UB(dst0, dst, stride);
+    ST_W2(dst0, 0, 1, dst, stride);
 }
 
 static void avc_biwgt_4x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -287,7 +287,7 @@  static void avc_biwgt_4x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     tmp1 >>= denom;
     CLIP_SH2_0_255(tmp0, tmp1);
     dst0 = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-    ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, stride);
+    ST_W4(dst0, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_biwgt_4x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -327,7 +327,7 @@  static void avc_biwgt_4x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SRA_4V(tmp0, tmp1, tmp2, tmp3, denom);
     CLIP_SH4_0_255(tmp0, tmp1, tmp2, tmp3);
     PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, dst0, dst1);
-    ST4x8_UB(dst0, dst1, dst, stride);
+    ST_W8(dst0, dst1, 0, 1, 2, 3, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_biwgt_8x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -365,7 +365,7 @@  static void avc_biwgt_8x4_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     SRA_4V(tmp0, tmp1, tmp2, tmp3, denom);
     CLIP_SH4_0_255(tmp0, tmp1, tmp2, tmp3);
     PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, dst0, dst1);
-    ST8x4_UB(dst0, dst1, dst, stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_biwgt_8x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -417,7 +417,7 @@  static void avc_biwgt_8x8_msa(uint8_t *src, uint8_t *dst, int32_t stride,
     CLIP_SH4_0_255(tmp4, tmp5, tmp6, tmp7);
     PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, dst0, dst1);
     PCKEV_B2_UB(tmp5, tmp4, tmp7, tmp6, dst2, dst3);
-    ST8x8_UB(dst0, dst1, dst2, dst3, dst, stride);
+    ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_biwgt_8x16_msa(uint8_t *src, uint8_t *dst, int32_t stride,
@@ -479,7 +479,7 @@  static void avc_biwgt_8x16_msa(uint8_t *src, uint8_t *dst, int32_t stride,
         CLIP_SH4_0_255(temp4, temp5, temp6, temp7);
         PCKEV_B4_UB(temp1, temp0, temp3, temp2, temp5, temp4, temp7, temp6,
                     dst0, dst1, dst2, dst3);
-        ST8x8_UB(dst0, dst1, dst2, dst3, dst, stride);
+        ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
         dst += 8 * stride;
     }
 }
@@ -955,18 +955,18 @@  static void avc_loopfilter_luma_intra_edge_ver_msa(uint8_t *data,
         ILVRL_H2_SH(tp3, tp2, tmp6, tmp7);
 
         src = data - 3;
-        ST4x4_UB(tmp3, tmp3, 0, 1, 2, 3, src, img_width);
-        ST2x4_UB(tmp2, 0, src + 4, img_width);
+        ST_W4(tmp3, 0, 1, 2, 3, src, img_width);
+        ST_H4(tmp2, 0, 1, 2, 3, src + 4, img_width);
         src += 4 * img_width;
-        ST4x4_UB(tmp4, tmp4, 0, 1, 2, 3, src, img_width);
-        ST2x4_UB(tmp2, 4, src + 4, img_width);
+        ST_W4(tmp4, 0, 1, 2, 3, src, img_width);
+        ST_H4(tmp2, 4, 5, 6, 7, src + 4, img_width);
         src += 4 * img_width;
 
-        ST4x4_UB(tmp6, tmp6, 0, 1, 2, 3, src, img_width);
-        ST2x4_UB(tmp5, 0, src + 4, img_width);
+        ST_W4(tmp6, 0, 1, 2, 3, src, img_width);
+        ST_H4(tmp5, 0, 1, 2, 3, src + 4, img_width);
         src += 4 * img_width;
-        ST4x4_UB(tmp7, tmp7, 0, 1, 2, 3, src, img_width);
-        ST2x4_UB(tmp5, 4, src + 4, img_width);
+        ST_W4(tmp7, 0, 1, 2, 3, src, img_width);
+        ST_H4(tmp5, 4, 5, 6, 7, src + 4, img_width);
     }
     }
 }
@@ -1274,9 +1274,9 @@  static void avc_loopfilter_cb_or_cr_intra_edge_ver_msa(uint8_t *data_cb_or_cr,
         tmp1 = (v8i16) __msa_ilvr_b((v16i8) q0_or_p0_org, (v16i8) p0_or_q0_org);
 
         data_cb_or_cr -= 1;
-        ST2x4_UB(tmp1, 0, data_cb_or_cr, img_width);
+        ST_H4(tmp1, 0, 1, 2, 3, data_cb_or_cr, img_width);
         data_cb_or_cr += 4 * img_width;
-        ST2x4_UB(tmp1, 4, data_cb_or_cr, img_width);
+        ST_H4(tmp1, 4, 5, 6, 7, data_cb_or_cr, img_width);
     }
 }
 
@@ -2110,9 +2110,9 @@  static void avc_loopfilter_cb_or_cr_inter_edge_ver_msa(uint8_t *data,
             q0_org = __msa_bmnz_v(q0_org, q0, is_less_than);
             tmp1 = (v8i16) __msa_ilvr_b((v16i8) q0_org, (v16i8) p0_org);
             src = data - 1;
-            ST2x4_UB(tmp1, 0, src, img_width);
+            ST_H4(tmp1, 0, 1, 2, 3, src, img_width);
             src += 4 * img_width;
-            ST2x4_UB(tmp1, 4, src, img_width);
+            ST_H4(tmp1, 4, 5, 6, 7, src, img_width);
         }
     }
 }
@@ -2136,7 +2136,7 @@  static void avc_h_loop_filter_chroma422_msa(uint8_t *src, int32_t stride,
         }
 
         AVC_LPF_H_CHROMA_422(src, stride, tc_val, alpha, beta, res);
-        ST2x4_UB(res, 0, (src - 1), stride);
+        ST_H4(res, 0, 1, 2, 3, (src - 1), stride);
         src += (4 * stride);
     }
 }
diff --git a/libavcodec/mips/h264idct_msa.c b/libavcodec/mips/h264idct_msa.c
index 1e1a5c8..7851bfd 100644
--- a/libavcodec/mips/h264idct_msa.c
+++ b/libavcodec/mips/h264idct_msa.c
@@ -237,9 +237,7 @@  static void avc_idct8_addblk_msa(uint8_t *dst, int16_t *src, int32_t dst_stride)
     CLIP_SH4_0_255(res4, res5, res6, res7);
     PCKEV_B4_SB(res1, res0, res3, res2, res5, res4, res7, res6,
                 dst0, dst1, dst2, dst3);
-    ST8x4_UB(dst0, dst1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST8x4_UB(dst2, dst3, dst, dst_stride);
+    ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride)
 }
 
 static void avc_idct8_dc_addblk_msa(uint8_t *dst, int16_t *src,
@@ -269,9 +267,7 @@  static void avc_idct8_dc_addblk_msa(uint8_t *dst, int16_t *src,
     CLIP_SH4_0_255(dst4_r, dst5_r, dst6_r, dst7_r);
     PCKEV_B4_SB(dst1_r, dst0_r, dst3_r, dst2_r, dst5_r, dst4_r, dst7_r, dst6_r,
                 dst0, dst1, dst2, dst3);
-    ST8x4_UB(dst0, dst1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST8x4_UB(dst2, dst3, dst, dst_stride);
+    ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride)
 }
 
 void ff_h264_idct_add_msa(uint8_t *dst, int16_t *src, int32_t dst_stride)
@@ -340,7 +336,7 @@  void ff_h264_idct4x4_addblk_dc_msa(uint8_t *dst, int16_t *src,
     ADD2(pred_r, input_dc, pred_l, input_dc, pred_r, pred_l);
     CLIP_SH2_0_255(pred_r, pred_l);
     out = __msa_pckev_b((v16i8) pred_l, (v16i8) pred_r);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 void ff_h264_idct8_dc_addblk_msa(uint8_t *dst, int16_t *src,
diff --git a/libavcodec/mips/h264qpel_msa.c b/libavcodec/mips/h264qpel_msa.c
index 9c779bd..df7e3e2 100644
--- a/libavcodec/mips/h264qpel_msa.c
+++ b/libavcodec/mips/h264qpel_msa.c
@@ -149,7 +149,7 @@  static void avc_luma_hv_qrt_4x4_msa(const uint8_t *src_x, const uint8_t *src_y,
 
     SAT_SH2_SH(out0, out1, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_luma_hv_qrt_8x8_msa(const uint8_t *src_x, const uint8_t *src_y,
@@ -220,7 +220,7 @@  static void avc_luma_hv_qrt_8x8_msa(const uint8_t *src_x, const uint8_t *src_y,
     SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
     out0 = PCKEV_XORI128_UB(tmp0, tmp1);
     out1 = PCKEV_XORI128_UB(tmp2, tmp3);
-    ST8x4_UB(out0, out1, dst, stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
     dst += (4 * stride);
 
     LD_SB4(src_y, stride, src_vt9, src_vt10, src_vt11, src_vt12);
@@ -256,8 +256,7 @@  static void avc_luma_hv_qrt_8x8_msa(const uint8_t *src_x, const uint8_t *src_y,
     SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
     out0 = PCKEV_XORI128_UB(tmp0, tmp1);
     out1 = PCKEV_XORI128_UB(tmp2, tmp3);
-    ST8x4_UB(out0, out1, dst, stride);
-    dst += (4 * stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_luma_hv_qrt_16x16_msa(const uint8_t *src_x,
@@ -337,7 +336,7 @@  static void avc_luma_hv_qrt_16x16_msa(const uint8_t *src_x,
             SAT_SH4_SH(out0, out1, out2, out3, 7);
             tmp0 = PCKEV_XORI128_UB(out0, out1);
             tmp1 = PCKEV_XORI128_UB(out2, out3);
-            ST8x4_UB(tmp0, tmp1, dst, stride);
+            ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, stride);
             dst += (4 * stride);
 
             src_vt0 = src_vt4;
@@ -419,7 +418,7 @@  static void avc_luma_hv_qrt_and_aver_dst_4x4_msa(const uint8_t *src_x,
     res = PCKEV_XORI128_UB(res0, res1);
     dst0 = __msa_aver_u_b(res, dst0);
 
-    ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, stride);
+    ST_W4(dst0, 0, 1, 2, 3, dst, stride);
 }
 
 static void avc_luma_hv_qrt_and_aver_dst_8x8_msa(const uint8_t *src_x,
@@ -498,7 +497,7 @@  static void avc_luma_hv_qrt_and_aver_dst_8x8_msa(const uint8_t *src_x,
     out0 = PCKEV_XORI128_UB(tmp0, tmp1);
     out1 = PCKEV_XORI128_UB(tmp2, tmp3);
     AVER_UB2_UB(out0, dst0, out1, dst1, dst0, dst1);
-    ST8x4_UB(dst0, dst1, dst, stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, stride);
     dst += (4 * stride);
 
     LD_SB4(src_y, stride, src_vt9, src_vt10, src_vt11, src_vt12);
@@ -539,8 +538,7 @@  static void avc_luma_hv_qrt_and_aver_dst_8x8_msa(const uint8_t *src_x,
     out0 = PCKEV_XORI128_UB(tmp0, tmp1);
     out1 = PCKEV_XORI128_UB(tmp2, tmp3);
     AVER_UB2_UB(out0, dst0, out1, dst1, dst0, dst1);
-    ST8x4_UB(dst0, dst1, dst, stride);
-    dst += (4 * stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, stride);
 }
 
 static void avc_luma_hv_qrt_and_aver_dst_16x16_msa(const uint8_t *src_x,
@@ -627,7 +625,7 @@  static void avc_luma_hv_qrt_and_aver_dst_16x16_msa(const uint8_t *src_x,
             tmp0 = PCKEV_XORI128_UB(out0, out1);
             tmp1 = PCKEV_XORI128_UB(out2, out3);
             AVER_UB2_UB(tmp0, dst0, tmp1, dst1, dst0, dst1);
-            ST8x4_UB(dst0, dst1, dst, stride);
+            ST_D4(dst0, dst1, 0, 1, 0, 1, dst, stride);
             dst += (4 * stride);
 
             src_vt0 = src_vt4;
@@ -723,7 +721,7 @@  void ff_avg_h264_qpel8_mc00_msa(uint8_t *dst, const uint8_t *src,
     AVER_UB4_UB(src0, dst0, src1, dst1, src2, dst2, src3, dst3, dst0, dst1,
                 dst2, dst3);
 
-    ST8x8_UB(dst0, dst1, dst2, dst3, dst, stride);
+    ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_avg_h264_qpel4_mc00_msa(uint8_t *dst, const uint8_t *src,
@@ -739,7 +737,7 @@  void ff_avg_h264_qpel4_mc00_msa(uint8_t *dst, const uint8_t *src,
 
     dst0 = __msa_aver_u_b(src0, dst0);
 
-    ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, stride);
+    ST_W4(dst0, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_put_h264_qpel16_mc10_msa(uint8_t *dst, const uint8_t *src,
@@ -930,7 +928,7 @@  void ff_put_h264_qpel8_mc10_msa(uint8_t *dst, const uint8_t *src,
     tmp2 = __msa_aver_s_b(tmp2, src4);
     tmp3 = __msa_aver_s_b(tmp3, src5);
     XORI_B4_128_SB(tmp0, tmp1, tmp2, tmp3);
-    ST8x8_UB(tmp0, tmp1, tmp2, tmp3, dst, stride);
+    ST_D8(tmp0, tmp1, tmp2, tmp3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_put_h264_qpel8_mc30_msa(uint8_t *dst, const uint8_t *src,
@@ -985,7 +983,7 @@  void ff_put_h264_qpel8_mc30_msa(uint8_t *dst, const uint8_t *src,
     tmp2 = __msa_aver_s_b(tmp2, src4);
     tmp3 = __msa_aver_s_b(tmp3, src5);
     XORI_B4_128_SB(tmp0, tmp1, tmp2, tmp3);
-    ST8x8_UB(tmp0, tmp1, tmp2, tmp3, dst, stride);
+    ST_D8(tmp0, tmp1, tmp2, tmp3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_put_h264_qpel4_mc10_msa(uint8_t *dst, const uint8_t *src,
@@ -1016,7 +1014,7 @@  void ff_put_h264_qpel4_mc10_msa(uint8_t *dst, const uint8_t *src,
     src0 = (v16i8) __msa_insve_d((v2i64) src0, 1, (v2i64) src1);
     res = __msa_aver_s_b(res, src0);
     res = (v16i8) __msa_xori_b((v16u8) res, 128);
-    ST4x4_UB(res, res, 0, 1, 2, 3, dst, stride);
+    ST_W4(res, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_put_h264_qpel4_mc30_msa(uint8_t *dst, const uint8_t *src,
@@ -1047,7 +1045,7 @@  void ff_put_h264_qpel4_mc30_msa(uint8_t *dst, const uint8_t *src,
     src0 = (v16i8) __msa_insve_d((v2i64) src0, 1, (v2i64) src1);
     res = __msa_aver_s_b(res, src0);
     res = (v16i8) __msa_xori_b((v16u8) res, 128);
-    ST4x4_UB(res, res, 0, 1, 2, 3, dst, stride);
+    ST_W4(res, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_put_h264_qpel16_mc20_msa(uint8_t *dst, const uint8_t *src,
@@ -1153,7 +1151,7 @@  void ff_put_h264_qpel8_mc20_msa(uint8_t *dst, const uint8_t *src,
     out1 = PCKEV_XORI128_UB(res2, res3);
     out2 = PCKEV_XORI128_UB(res4, res5);
     out3 = PCKEV_XORI128_UB(res6, res7);
-    ST8x8_UB(out0, out1, out2, out3, dst, stride);
+    ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_put_h264_qpel4_mc20_msa(uint8_t *dst, const uint8_t *src,
@@ -1178,7 +1176,7 @@  void ff_put_h264_qpel4_mc20_msa(uint8_t *dst, const uint8_t *src,
     SRARI_H2_SH(res0, res1, 5);
     SAT_SH2_SH(res0, res1, 7);
     out = PCKEV_XORI128_UB(res0, res1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_put_h264_qpel16_mc01_msa(uint8_t *dst, const uint8_t *src,
@@ -1378,7 +1376,7 @@  void ff_put_h264_qpel8_mc01_msa(uint8_t *dst, const uint8_t *src,
     out2 = __msa_aver_s_b(out2, tmp2);
     out3 = __msa_aver_s_b(out3, tmp3);
     XORI_B4_128_SB(out0, out1, out2, out3);
-    ST8x8_UB(out0, out1, out2, out3, dst, stride);
+    ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_put_h264_qpel8_mc03_msa(uint8_t *dst, const uint8_t *src,
@@ -1431,7 +1429,7 @@  void ff_put_h264_qpel8_mc03_msa(uint8_t *dst, const uint8_t *src,
     out2 = __msa_aver_s_b(out2, tmp2);
     out3 = __msa_aver_s_b(out3, tmp3);
     XORI_B4_128_SB(out0, out1, out2, out3);
-    ST8x8_UB(out0, out1, out2, out3, dst, stride);
+    ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_put_h264_qpel4_mc01_msa(uint8_t *dst, const uint8_t *src,
@@ -1472,7 +1470,7 @@  void ff_put_h264_qpel4_mc01_msa(uint8_t *dst, const uint8_t *src,
     src54_r = (v16i8) __msa_insve_w((v4i32) src4, 1, (v4i32) src5);
     src32_r = (v16i8) __msa_insve_d((v2i64) src32_r, 1, (v2i64) src54_r);
     out = __msa_aver_u_b(out, (v16u8) src32_r);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_put_h264_qpel4_mc03_msa(uint8_t *dst, const uint8_t *src,
@@ -1513,7 +1511,7 @@  void ff_put_h264_qpel4_mc03_msa(uint8_t *dst, const uint8_t *src,
     src54_r = (v16i8) __msa_insve_w((v4i32) src5, 1, (v4i32) src6);
     src32_r = (v16i8) __msa_insve_d((v2i64) src32_r, 1, (v2i64) src54_r);
     out = __msa_aver_u_b(out, (v16u8) src32_r);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_put_h264_qpel16_mc11_msa(uint8_t *dst, const uint8_t *src,
@@ -1691,7 +1689,7 @@  void ff_put_h264_qpel16_mc21_msa(uint8_t *dst, const uint8_t *src,
 
             out0 = PCKEV_XORI128_UB(dst0, dst1);
             out1 = PCKEV_XORI128_UB(dst2, dst3);
-            ST8x4_UB(out0, out1, dst, stride);
+            ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
             dst += (4 * stride);
 
             hz_out0 = hz_out4;
@@ -1804,7 +1802,7 @@  void ff_put_h264_qpel16_mc23_msa(uint8_t *dst, const uint8_t *src,
 
             out0 = PCKEV_XORI128_UB(dst0, dst1);
             out1 = PCKEV_XORI128_UB(dst2, dst3);
-            ST8x4_UB(out0, out1, dst, stride);
+            ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
             dst += (4 * stride);
 
             hz_out0 = hz_out4;
@@ -1905,7 +1903,7 @@  void ff_put_h264_qpel8_mc21_msa(uint8_t *dst, const uint8_t *src,
 
     out0 = PCKEV_XORI128_UB(dst0, dst1);
     out1 = PCKEV_XORI128_UB(dst2, dst3);
-    ST8x4_UB(out0, out1, dst, stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
     dst += (4 * stride);
 
     LD_SB4(src, stride, src9, src10, src11, src12);
@@ -1951,7 +1949,7 @@  void ff_put_h264_qpel8_mc21_msa(uint8_t *dst, const uint8_t *src,
 
     out0 = PCKEV_XORI128_UB(dst0, dst1);
     out1 = PCKEV_XORI128_UB(dst2, dst3);
-    ST8x4_UB(out0, out1, dst, stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_put_h264_qpel8_mc23_msa(uint8_t *dst, const uint8_t *src,
@@ -2040,7 +2038,7 @@  void ff_put_h264_qpel8_mc23_msa(uint8_t *dst, const uint8_t *src,
 
     out0 = PCKEV_XORI128_UB(dst0, dst1);
     out1 = PCKEV_XORI128_UB(dst2, dst3);
-    ST8x4_UB(out0, out1, dst, stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
     dst += (4 * stride);
 
     LD_SB4(src, stride, src9, src10, src11, src12);
@@ -2086,7 +2084,7 @@  void ff_put_h264_qpel8_mc23_msa(uint8_t *dst, const uint8_t *src,
 
     out0 = PCKEV_XORI128_UB(dst0, dst1);
     out1 = PCKEV_XORI128_UB(dst2, dst3);
-    ST8x4_UB(out0, out1, dst, stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_put_h264_qpel4_mc21_msa(uint8_t *dst, const uint8_t *src,
@@ -2150,7 +2148,7 @@  void ff_put_h264_qpel4_mc21_msa(uint8_t *dst, const uint8_t *src,
     dst1 = __msa_aver_s_h(dst1, hz_out4);
 
     res = PCKEV_XORI128_UB(dst0, dst1);
-    ST4x4_UB(res, res, 0, 1, 2, 3, dst, stride);
+    ST_W4(res, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_put_h264_qpel4_mc23_msa(uint8_t *dst, const uint8_t *src,
@@ -2215,7 +2213,7 @@  void ff_put_h264_qpel4_mc23_msa(uint8_t *dst, const uint8_t *src,
     dst1 = __msa_aver_s_h(dst1, hz_out1);
 
     res = PCKEV_XORI128_UB(dst0, dst1);
-    ST4x4_UB(res, res, 0, 1, 2, 3, dst, stride);
+    ST_W4(res, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_put_h264_qpel16_mc02_msa(uint8_t *dst, const uint8_t *src,
@@ -2332,7 +2330,7 @@  void ff_put_h264_qpel8_mc02_msa(uint8_t *dst, const uint8_t *src,
     out1 = PCKEV_XORI128_UB(out2_r, out3_r);
     out2 = PCKEV_XORI128_UB(out4_r, out5_r);
     out3 = PCKEV_XORI128_UB(out6_r, out7_r);
-    ST8x8_UB(out0, out1, out2, out3, dst, stride);
+    ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_put_h264_qpel4_mc02_msa(uint8_t *dst, const uint8_t *src,
@@ -2369,7 +2367,7 @@  void ff_put_h264_qpel4_mc02_msa(uint8_t *dst, const uint8_t *src,
     SRARI_H2_SH(out10, out32, 5);
     SAT_SH2_SH(out10, out32, 7);
     out = PCKEV_XORI128_UB(out10, out32);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_put_h264_qpel16_mc12_msa(uint8_t *dst, const uint8_t *src,
@@ -2601,7 +2599,7 @@  void ff_put_h264_qpel8_mc12_msa(uint8_t *dst, const uint8_t *src,
         dst0 = __msa_aver_s_h(dst2, dst0);
         dst1 = __msa_aver_s_h(dst3, dst1);
         out = PCKEV_XORI128_UB(dst0, dst1);
-        ST8x2_UB(out, dst, stride);
+        ST_D2(out, 0, 1, dst, stride);
         dst += (2 * stride);
 
         src0 = src2;
@@ -2677,7 +2675,7 @@  void ff_put_h264_qpel8_mc32_msa(uint8_t *dst, const uint8_t *src,
         dst0 = __msa_aver_s_h(dst2, dst0);
         dst1 = __msa_aver_s_h(dst3, dst1);
         out = PCKEV_XORI128_UB(dst0, dst1);
-        ST8x2_UB(out, dst, stride);
+        ST_D2(out, 0, 1, dst, stride);
         dst += (2 * stride);
 
         src0 = src2;
@@ -2777,7 +2775,7 @@  void ff_put_h264_qpel4_mc12_msa(uint8_t *dst, const uint8_t *src,
 
     PCKEV_H2_SH(hz_res1, hz_res0, hz_res3, hz_res2, dst0, dst2);
     out = PCKEV_XORI128_UB(dst0, dst2);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_put_h264_qpel4_mc32_msa(uint8_t *dst, const uint8_t *src,
@@ -2873,7 +2871,7 @@  void ff_put_h264_qpel4_mc32_msa(uint8_t *dst, const uint8_t *src,
 
     PCKEV_H2_SH(hz_res1, hz_res0, hz_res3, hz_res2, dst0, dst2);
     out = PCKEV_XORI128_UB(dst0, dst2);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_put_h264_qpel16_mc22_msa(uint8_t *dst, const uint8_t *src,
@@ -2961,7 +2959,7 @@  void ff_put_h264_qpel16_mc22_msa(uint8_t *dst, const uint8_t *src,
 
             out0 = PCKEV_XORI128_UB(dst0, dst1);
             out1 = PCKEV_XORI128_UB(dst2, dst3);
-            ST8x4_UB(out0, out1, dst, stride);
+            ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
             dst += (4 * stride);
 
             hz_out0 = hz_out4;
@@ -3049,7 +3047,7 @@  void ff_put_h264_qpel8_mc22_msa(uint8_t *dst, const uint8_t *src,
     dst3 = __msa_pckev_h((v8i16) tmp1, (v8i16) tmp0);
     out0 = PCKEV_XORI128_UB(dst0, dst1);
     out1 = PCKEV_XORI128_UB(dst2, dst3);
-    ST8x4_UB(out0, out1, dst, stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
     dst += (4 * stride);
 
     LD_SB4(src, stride, src0, src1, src2, src3);
@@ -3086,7 +3084,7 @@  void ff_put_h264_qpel8_mc22_msa(uint8_t *dst, const uint8_t *src,
     dst3 = __msa_pckev_h((v8i16) tmp1, (v8i16) tmp0);
     out0 = PCKEV_XORI128_UB(dst0, dst1);
     out1 = PCKEV_XORI128_UB(dst2, dst3);
-    ST8x4_UB(out0, out1, dst, stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_put_h264_qpel4_mc22_msa(uint8_t *dst, const uint8_t *src,
@@ -3141,7 +3139,7 @@  void ff_put_h264_qpel4_mc22_msa(uint8_t *dst, const uint8_t *src,
                           filt2);
     dst1 = __msa_pckev_h((v8i16) tmp1, (v8i16) tmp0);
     res = PCKEV_XORI128_UB(dst0, dst1);
-    ST4x4_UB(res, res, 0, 1, 2, 3, dst, stride);
+    ST_W4(res, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_avg_h264_qpel16_mc10_msa(uint8_t *dst, const uint8_t *src,
@@ -3350,7 +3348,7 @@  void ff_avg_h264_qpel8_mc10_msa(uint8_t *dst, const uint8_t *src,
     INSERT_D2_UB(tp2, tp3, dst3);
     AVER_UB2_UB(tmp0, dst0, tmp1, dst1, dst0, dst1);
     AVER_UB2_UB(tmp2, dst2, tmp3, dst3, dst2, dst3);
-    ST8x8_UB(dst0, dst1, dst2, dst3, dst, stride);
+    ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_avg_h264_qpel8_mc30_msa(uint8_t *dst, const uint8_t *src,
@@ -3415,7 +3413,7 @@  void ff_avg_h264_qpel8_mc30_msa(uint8_t *dst, const uint8_t *src,
     INSERT_D2_UB(tp2, tp3, dst3);
     AVER_UB2_UB(tmp0, dst0, tmp1, dst1, dst0, dst1);
     AVER_UB2_UB(tmp2, dst2, tmp3, dst3, dst2, dst3);
-    ST8x8_UB(dst0, dst1, dst2, dst3, dst, stride);
+    ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_avg_h264_qpel4_mc10_msa(uint8_t *dst, const uint8_t *src,
@@ -3451,7 +3449,7 @@  void ff_avg_h264_qpel4_mc10_msa(uint8_t *dst, const uint8_t *src,
     LW4(dst, stride, tp0, tp1, tp2, tp3);
     INSERT_W4_UB(tp0, tp1, tp2, tp3, dst0);
     dst0 = __msa_aver_u_b((v16u8) res, dst0);
-    ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, stride);
+    ST_W4(dst0, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_avg_h264_qpel4_mc30_msa(uint8_t *dst, const uint8_t *src,
@@ -3487,7 +3485,7 @@  void ff_avg_h264_qpel4_mc30_msa(uint8_t *dst, const uint8_t *src,
     LW4(dst, stride, tp0, tp1, tp2, tp3);
     INSERT_W4_UB(tp0, tp1, tp2, tp3, dst0);
     dst0 = __msa_aver_u_b((v16u8) res, dst0);
-    ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, stride);
+    ST_W4(dst0, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_avg_h264_qpel16_mc20_msa(uint8_t *dst, const uint8_t *src,
@@ -3608,7 +3606,7 @@  void ff_avg_h264_qpel8_mc20_msa(uint8_t *dst, const uint8_t *src,
     INSERT_D2_UB(tp2, tp3, out7);
     AVER_UB2_UB(out0, out2, out1, out3, out0, out1);
     AVER_UB2_UB(out4, out6, out5, out7, out4, out5);
-    ST8x8_UB(out0, out1, out4, out5, dst, stride);
+    ST_D8(out0, out1, out4, out5, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_avg_h264_qpel4_mc20_msa(uint8_t *dst, const uint8_t *src,
@@ -3637,7 +3635,7 @@  void ff_avg_h264_qpel4_mc20_msa(uint8_t *dst, const uint8_t *src,
     LW4(dst, stride, tp0, tp1, tp2, tp3);
     INSERT_W4_UB(tp0, tp1, tp2, tp3, dst0);
     res = __msa_aver_u_b(res, dst0);
-    ST4x4_UB(res, res, 0, 1, 2, 3, dst, stride);
+    ST_W4(res, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_avg_h264_qpel16_mc01_msa(uint8_t *dst, const uint8_t *src,
@@ -3856,7 +3854,7 @@  void ff_avg_h264_qpel8_mc01_msa(uint8_t *dst, const uint8_t *src,
     XORI_B4_128_SB(out0, out1, out2, out3);
     AVER_UB4_UB(out0, dst0, out1, dst1, out2, dst2, out3, dst3, dst0, dst1,
                 dst2, dst3);
-    ST8x8_UB(dst0, dst1, dst2, dst3, dst, stride);
+    ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_avg_h264_qpel8_mc03_msa(uint8_t *dst, const uint8_t *src,
@@ -3922,7 +3920,7 @@  void ff_avg_h264_qpel8_mc03_msa(uint8_t *dst, const uint8_t *src,
     XORI_B4_128_SB(out0, out1, out2, out3);
     AVER_UB4_UB(out0, dst0, out1, dst1, out2, dst2, out3, dst3, dst0, dst1,
                 dst2, dst3);
-    ST8x8_UB(dst0, dst1, dst2, dst3, dst, stride);
+    ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_avg_h264_qpel4_mc01_msa(uint8_t *dst, const uint8_t *src,
@@ -3967,7 +3965,7 @@  void ff_avg_h264_qpel4_mc01_msa(uint8_t *dst, const uint8_t *src,
     res = PCKEV_XORI128_UB(out10, out32);
     res = __msa_aver_u_b(res, (v16u8) src32_r);
     dst0 = __msa_aver_u_b(res, dst0);
-    ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, stride);
+    ST_W4(dst0, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_avg_h264_qpel4_mc03_msa(uint8_t *dst, const uint8_t *src,
@@ -4013,7 +4011,7 @@  void ff_avg_h264_qpel4_mc03_msa(uint8_t *dst, const uint8_t *src,
     src32_r = (v16i8) __msa_insve_d((v2i64) src32_r, 1, (v2i64) src54_r);
     res = __msa_aver_u_b(res, (v16u8) src32_r);
     dst0 = __msa_aver_u_b(res, dst0);
-    ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, stride);
+    ST_W4(dst0, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_avg_h264_qpel16_mc11_msa(uint8_t *dst, const uint8_t *src,
@@ -4196,7 +4194,7 @@  void ff_avg_h264_qpel16_mc21_msa(uint8_t *dst, const uint8_t *src,
 
             out0 = PCKEV_XORI128_UB(tmp0, tmp1);
             dst0 = __msa_aver_u_b(out0, dst0);
-            ST8x2_UB(dst0, dst, stride);
+            ST_D2(dst0, 0, 1, dst, stride);
             dst += (2 * stride);
 
             LD_SB2(src, stride, src7, src8);
@@ -4232,7 +4230,7 @@  void ff_avg_h264_qpel16_mc21_msa(uint8_t *dst, const uint8_t *src,
 
             out1 = PCKEV_XORI128_UB(tmp2, tmp3);
             dst1 = __msa_aver_u_b(out1, dst1);
-            ST8x2_UB(dst1, dst, stride);
+            ST_D2(dst1, 0, 1, dst, stride);
             dst += (2 * stride);
 
             hz_out0 = hz_out4;
@@ -4326,7 +4324,7 @@  void ff_avg_h264_qpel16_mc23_msa(uint8_t *dst, const uint8_t *src,
             INSERT_D2_UB(tp0, tp1, dst0);
             out0 = PCKEV_XORI128_UB(tmp0, tmp1);
             dst0 = __msa_aver_u_b(out0, dst0);
-            ST8x2_UB(dst0, dst, stride);
+            ST_D2(dst0, 0, 1, dst, stride);
             dst += (2 * stride);
 
             LD_SB2(src, stride, src7, src8);
@@ -4361,7 +4359,7 @@  void ff_avg_h264_qpel16_mc23_msa(uint8_t *dst, const uint8_t *src,
             INSERT_D2_UB(tp2, tp3, dst1);
             out1 = PCKEV_XORI128_UB(tmp2, tmp3);
             dst1 = __msa_aver_u_b(out1, dst1);
-            ST8x2_UB(dst1, dst, stride);
+            ST_D2(dst1, 0, 1, dst, stride);
             dst += (2 * stride);
 
             hz_out0 = hz_out4;
@@ -4468,7 +4466,7 @@  void ff_avg_h264_qpel8_mc21_msa(uint8_t *dst, const uint8_t *src,
     out0 = PCKEV_XORI128_UB(tmp0, tmp1);
     out1 = PCKEV_XORI128_UB(tmp2, tmp3);
     AVER_UB2_UB(out0, dst0, out1, dst1, dst0, dst1);
-    ST8x4_UB(dst0, dst1, dst, stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, stride);
     dst += (4 * stride);
 
     LD_SB4(src, stride, src9, src10, src11, src12);
@@ -4519,7 +4517,7 @@  void ff_avg_h264_qpel8_mc21_msa(uint8_t *dst, const uint8_t *src,
     out0 = PCKEV_XORI128_UB(tmp0, tmp1);
     out1 = PCKEV_XORI128_UB(tmp2, tmp3);
     AVER_UB2_UB(out0, dst0, out1, dst1, dst0, dst1);
-    ST8x4_UB(dst0, dst1, dst, stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_avg_h264_qpel8_mc23_msa(uint8_t *dst, const uint8_t *src,
@@ -4614,7 +4612,7 @@  void ff_avg_h264_qpel8_mc23_msa(uint8_t *dst, const uint8_t *src,
     out0 = PCKEV_XORI128_UB(tmp0, tmp1);
     out1 = PCKEV_XORI128_UB(tmp2, tmp3);
     AVER_UB2_UB(out0, dst0, out1, dst1, dst0, dst1);
-    ST8x4_UB(dst0, dst1, dst, stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, stride);
     dst += (4 * stride);
 
     LD_SB4(src, stride, src9, src10, src11, src12);
@@ -4665,7 +4663,7 @@  void ff_avg_h264_qpel8_mc23_msa(uint8_t *dst, const uint8_t *src,
     out0 = PCKEV_XORI128_UB(tmp0, tmp1);
     out1 = PCKEV_XORI128_UB(tmp2, tmp3);
     AVER_UB2_UB(out0, dst0, out1, dst1, dst0, dst1);
-    ST8x4_UB(dst0, dst1, dst, stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_avg_h264_qpel4_mc21_msa(uint8_t *dst, const uint8_t *src,
@@ -4732,7 +4730,7 @@  void ff_avg_h264_qpel4_mc21_msa(uint8_t *dst, const uint8_t *src,
     INSERT_W4_UB(tp0, tp1, tp2, tp3, out);
     res = PCKEV_XORI128_UB(dst0, dst1);
     res = __msa_aver_u_b(res, out);
-    ST4x4_UB(res, res, 0, 1, 2, 3, dst, stride);
+    ST_W4(res, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_avg_h264_qpel4_mc23_msa(uint8_t *dst, const uint8_t *src,
@@ -4800,7 +4798,7 @@  void ff_avg_h264_qpel4_mc23_msa(uint8_t *dst, const uint8_t *src,
     INSERT_W4_UB(tp0, tp1, tp2, tp3, out);
     res = PCKEV_XORI128_UB(dst0, dst1);
     res = __msa_aver_u_b(res, out);
-    ST4x4_UB(res, res, 0, 1, 2, 3, dst, stride);
+    ST_W4(res, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_avg_h264_qpel16_mc02_msa(uint8_t *dst, const uint8_t *src,
@@ -4936,7 +4934,7 @@  void ff_avg_h264_qpel8_mc02_msa(uint8_t *dst, const uint8_t *src,
     out3 = PCKEV_XORI128_UB(out6_r, out7_r);
     AVER_UB4_UB(out0, dst0, out1, dst1, out2, dst2, out3, dst3, dst0, dst1,
                 dst2, dst3);
-    ST8x8_UB(dst0, dst1, dst2, dst3, dst, stride);
+    ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_avg_h264_qpel4_mc02_msa(uint8_t *dst, const uint8_t *src,
@@ -4977,7 +4975,7 @@  void ff_avg_h264_qpel4_mc02_msa(uint8_t *dst, const uint8_t *src,
     INSERT_W4_UB(tp0, tp1, tp2, tp3, dst0);
     res = PCKEV_XORI128_UB(out10, out32);
     dst0 = __msa_aver_u_b(res, dst0);
-    ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, stride);
+    ST_W4(dst0, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_avg_h264_qpel16_mc12_msa(uint8_t *dst, const uint8_t *src,
@@ -5217,7 +5215,7 @@  void ff_avg_h264_qpel8_mc12_msa(uint8_t *dst, const uint8_t *src,
         tmp1 = __msa_aver_s_h(tmp3, tmp1);
         out = PCKEV_XORI128_UB(tmp0, tmp1);
         out = __msa_aver_u_b(out, dst0);
-        ST8x2_UB(out, dst, stride);
+        ST_D2(out, 0, 1, dst, stride);
         dst += (2 * stride);
 
         src0 = src2;
@@ -5297,7 +5295,7 @@  void ff_avg_h264_qpel8_mc32_msa(uint8_t *dst, const uint8_t *src,
         tmp1 = __msa_aver_s_h(tmp3, tmp1);
         out = PCKEV_XORI128_UB(tmp0, tmp1);
         out = __msa_aver_u_b(out, dst0);
-        ST8x2_UB(out, dst, stride);
+        ST_D2(out, 0, 1, dst, stride);
         dst += (2 * stride);
 
         src0 = src2;
@@ -5401,7 +5399,7 @@  void ff_avg_h264_qpel4_mc12_msa(uint8_t *dst, const uint8_t *src,
     PCKEV_H2_SH(hz_res1, hz_res0, hz_res3, hz_res2, dst0, dst2);
     out = PCKEV_XORI128_UB(dst0, dst2);
     out = __msa_aver_u_b(out, dstv);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_avg_h264_qpel4_mc32_msa(uint8_t *dst, const uint8_t *src,
@@ -5500,7 +5498,7 @@  void ff_avg_h264_qpel4_mc32_msa(uint8_t *dst, const uint8_t *src,
     PCKEV_H2_SH(hz_res1, hz_res0, hz_res3, hz_res2, dst0, dst2);
     out = PCKEV_XORI128_UB(dst0, dst2);
     out = __msa_aver_u_b(out, dstv);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, stride);
+    ST_W4(out, 0, 1, 2, 3, dst, stride);
 }
 
 void ff_avg_h264_qpel16_mc22_msa(uint8_t *dst, const uint8_t *src,
@@ -5592,7 +5590,7 @@  void ff_avg_h264_qpel16_mc22_msa(uint8_t *dst, const uint8_t *src,
             out0 = PCKEV_XORI128_UB(res0, res1);
             out1 = PCKEV_XORI128_UB(res2, res3);
             AVER_UB2_UB(out0, dst0, out1, dst1, out0, out1);
-            ST8x4_UB(out0, out1, dst, stride);
+            ST_D4(out0, out1, 0, 1, 0, 1, dst, stride);
             dst += (4 * stride);
 
             hz_out0 = hz_out4;
@@ -5685,7 +5683,7 @@  void ff_avg_h264_qpel8_mc22_msa(uint8_t *dst, const uint8_t *src,
     out0 = PCKEV_XORI128_UB(res0, res1);
     out1 = PCKEV_XORI128_UB(res2, res3);
     AVER_UB2_UB(out0, dst0, out1, dst1, dst0, dst1);
-    ST8x4_UB(dst0, dst1, dst, stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, stride);
     dst += (4 * stride);
 
     LD_SB4(src, stride, src0, src1, src2, src3);
@@ -5726,7 +5724,7 @@  void ff_avg_h264_qpel8_mc22_msa(uint8_t *dst, const uint8_t *src,
     out0 = PCKEV_XORI128_UB(res0, res1);
     out1 = PCKEV_XORI128_UB(res2, res3);
     AVER_UB2_UB(out0, dst0, out1, dst1, dst0, dst1);
-    ST8x4_UB(dst0, dst1, dst, stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, stride);
 }
 
 void ff_avg_h264_qpel4_mc22_msa(uint8_t *dst, const uint8_t *src,
@@ -5785,5 +5783,5 @@  void ff_avg_h264_qpel4_mc22_msa(uint8_t *dst, const uint8_t *src,
     INSERT_W4_UB(tp0, tp1, tp2, tp3, dst0);
     res = PCKEV_XORI128_UB(res0, res1);
     res = __msa_aver_u_b(res, dst0);
-    ST4x4_UB(res, res, 0, 1, 2, 3, dst, stride);
+    ST_W4(res, 0, 1, 2, 3, dst, stride);
 }
diff --git a/libavcodec/mips/hevc_idct_msa.c b/libavcodec/mips/hevc_idct_msa.c
index 0943119..b14aec9 100644
--- a/libavcodec/mips/hevc_idct_msa.c
+++ b/libavcodec/mips/hevc_idct_msa.c
@@ -727,7 +727,7 @@  static void hevc_addblk_4x4_msa(int16_t *coeffs, uint8_t *dst, int32_t stride)
     ADD2(dst_r0, in0, dst_l0, in1, dst_r0, dst_l0);
     CLIP_SH2_0_255(dst_r0, dst_l0);
     dst_vec = (v4i32) __msa_pckev_b((v16i8) dst_l0, (v16i8) dst_r0);
-    ST4x4_UB(dst_vec, dst_vec, 0, 1, 2, 3, dst, stride);
+    ST_W4(dst_vec, 0, 1, 2, 3, dst, stride);
 }
 
 static void hevc_addblk_8x8_msa(int16_t *coeffs, uint8_t *dst, int32_t stride)
@@ -752,8 +752,7 @@  static void hevc_addblk_8x8_msa(int16_t *coeffs, uint8_t *dst, int32_t stride)
          dst_r0, dst_l0, dst_r1, dst_l1);
     CLIP_SH4_0_255(dst_r0, dst_l0, dst_r1, dst_l1);
     PCKEV_B2_SH(dst_l0, dst_r0, dst_l1, dst_r1, dst_r0, dst_r1);
-    ST8x4_UB(dst_r0, dst_r1, dst, stride);
-    dst += (4 * stride);
+    ST_D4(dst_r0, dst_r1, 0, 1, 0, 1, dst, stride);
 
     LD4(temp_dst, stride, dst0, dst1, dst2, dst3);
     INSERT_D2_SD(dst0, dst1, dst_vec0);
@@ -764,7 +763,7 @@  static void hevc_addblk_8x8_msa(int16_t *coeffs, uint8_t *dst, int32_t stride)
          dst_r0, dst_l0, dst_r1, dst_l1);
     CLIP_SH4_0_255(dst_r0, dst_l0, dst_r1, dst_l1);
     PCKEV_B2_SH(dst_l0, dst_r0, dst_l1, dst_r1, dst_r0, dst_r1);
-    ST8x4_UB(dst_r0, dst_r1, dst, stride);
+    ST_D4(dst_r0, dst_r1, 0, 1, 0, 1, dst + 4 * stride, stride);
 }
 
 static void hevc_addblk_16x16_msa(int16_t *coeffs, uint8_t *dst, int32_t stride)
diff --git a/libavcodec/mips/hevc_lpf_sao_msa.c b/libavcodec/mips/hevc_lpf_sao_msa.c
index adcafde..ac21806 100644
--- a/libavcodec/mips/hevc_lpf_sao_msa.c
+++ b/libavcodec/mips/hevc_lpf_sao_msa.c
@@ -199,11 +199,9 @@  static void hevc_loopfilter_luma_hor_msa(uint8_t *src, int32_t stride,
             dst_val0 = __msa_copy_u_d((v2i64) dst2, 0);
             dst_val1 = __msa_copy_u_d((v2i64) dst2, 1);
 
-            ST8x4_UB(dst0, dst1, p2, stride);
-            p2 += (4 * stride);
-            SD(dst_val0, p2);
-            p2 += stride;
-            SD(dst_val1, p2);
+            ST_D4(dst0, dst1, 0, 1, 0, 1, p2, stride);
+            SD(dst_val0, p2 + 4 * stride);
+            SD(dst_val1, p2 + 5 * stride);
             /* strong filter ends */
         } else if (flag0 == flag1) { /* weak only */
             /* weak filter */
@@ -288,7 +286,7 @@  static void hevc_loopfilter_luma_hor_msa(uint8_t *src, int32_t stride,
             dst1 = __msa_bmz_v(dst1, dst3, (v16u8) cmp3);
 
             p2 += stride;
-            ST8x4_UB(dst0, dst1, p2, stride);
+            ST_D4(dst0, dst1, 0, 1, 0, 1, p2, stride);
             /* weak filter ends */
         } else { /* strong + weak */
             /* strong filter */
@@ -442,11 +440,9 @@  static void hevc_loopfilter_luma_hor_msa(uint8_t *src, int32_t stride,
             dst_val0 = __msa_copy_u_d((v2i64) dst2, 0);
             dst_val1 = __msa_copy_u_d((v2i64) dst2, 1);
 
-            ST8x4_UB(dst0, dst1, p2, stride);
-            p2 += (4 * stride);
-            SD(dst_val0, p2);
-            p2 += stride;
-            SD(dst_val1, p2);
+            ST_D4(dst0, dst1, 0, 1, 0, 1, p2, stride);
+            SD(dst_val0, p2 + 4 * stride);
+            SD(dst_val1, p2 + 5 * stride);
         }
     }
 }
@@ -976,7 +972,7 @@  static void hevc_loopfilter_chroma_hor_msa(uint8_t *src, int32_t stride,
         temp1 = (v8i16) __msa_bmnz_v((v16u8) temp1, (v16u8) q0, (v16u8) tc_pos);
 
         temp0 = (v8i16) __msa_pckev_b((v16i8) temp1, (v16i8) temp0);
-        ST8x2_UB(temp0, p0_ptr, stride);
+        ST_D2(temp0, 0, 1, p0_ptr, stride);
     }
 }
 
@@ -1037,9 +1033,7 @@  static void hevc_loopfilter_chroma_ver_msa(uint8_t *src, int32_t stride,
         temp0 = (v8i16) __msa_ilvev_b((v16i8) temp1, (v16i8) temp0);
 
         src += 1;
-        ST2x4_UB(temp0, 0, src, stride);
-        src += (4 * stride);
-        ST2x4_UB(temp0, 4, src, stride);
+        ST_H8(temp0, 0, 1, 2, 3, 4, 5, 6, 7, src, stride);
     }
 }
 
@@ -1087,7 +1081,7 @@  static void hevc_sao_band_filter_4width_msa(uint8_t *dst, int32_t dst_stride,
         LD_UB4(src, src_stride, src0, src1, src2, src3);
 
         /* store results */
-        ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(dst0, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 
@@ -1102,7 +1096,7 @@  static void hevc_sao_band_filter_4width_msa(uint8_t *dst, int32_t dst_stride,
     dst0 = (v16i8) __msa_xori_b((v16u8) dst0, 128);
 
     /* store results */
-    ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(dst0, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void hevc_sao_band_filter_8width_msa(uint8_t *dst, int32_t dst_stride,
@@ -1153,7 +1147,7 @@  static void hevc_sao_band_filter_8width_msa(uint8_t *dst, int32_t dst_stride,
         XORI_B2_128_SB(dst0, dst1);
 
         /* store results */
-        ST8x4_UB(dst0, dst1, dst, dst_stride);
+        ST_D4(dst0, dst1, 0, 1, 0, 1, dst, dst_stride);
         dst += dst_stride << 2;
     }
 
@@ -1173,7 +1167,7 @@  static void hevc_sao_band_filter_8width_msa(uint8_t *dst, int32_t dst_stride,
     XORI_B2_128_SB(dst0, dst1);
 
     /* store results */
-    ST8x4_UB(dst0, dst1, dst, dst_stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hevc_sao_band_filter_16multiple_msa(uint8_t *dst,
diff --git a/libavcodec/mips/hevc_mc_bi_msa.c b/libavcodec/mips/hevc_mc_bi_msa.c
index b555517..34613c8 100644
--- a/libavcodec/mips/hevc_mc_bi_msa.c
+++ b/libavcodec/mips/hevc_mc_bi_msa.c
@@ -86,7 +86,7 @@  static void hevc_bi_copy_4w_msa(uint8_t *src0_ptr,
         dst0 = CLIP_SH_0_255_MAX_SATU(dst0);
 
         dst0 = (v8i16) __msa_pckev_b((v16i8) dst0, (v16i8) dst0);
-        ST4x2_UB(dst0, dst, dst_stride);
+        ST_W2(dst0, 0, 1, dst, dst_stride);
     } else if (4 == height) {
         LW4(src0_ptr, src_stride, tp0, tp1, tp2, tp3);
         INSERT_W4_SB(tp0, tp1, tp2, tp3, src0);
@@ -97,7 +97,7 @@  static void hevc_bi_copy_4w_msa(uint8_t *src0_ptr,
         SLLI_2V(dst0, dst1, 6);
         HEVC_BI_RND_CLIP2_MAX_SATU(in0, in1, dst0, dst1, 7, dst0, dst1);
         dst0 = (v8i16) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-        ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(dst0, 0, 1, 2, 3, dst, dst_stride);
     } else if (0 == height % 8) {
         for (loop_cnt = (height >> 3); loop_cnt--;) {
             LW4(src0_ptr, src_stride, tp0, tp1, tp2, tp3);
@@ -120,7 +120,7 @@  static void hevc_bi_copy_4w_msa(uint8_t *src0_ptr,
             HEVC_BI_RND_CLIP4_MAX_SATU(in0, in1, in2, in3, dst0, dst1, dst2,
                                        dst3, 7, dst0, dst1, dst2, dst3);
             PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
-            ST4x8_UB(dst0, dst1, dst, dst_stride);
+            ST_W8(dst0, dst1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
             dst += (8 * dst_stride);
         }
     }
@@ -165,9 +165,15 @@  static void hevc_bi_copy_6w_msa(uint8_t *src0_ptr,
                                    7, dst4, dst5, dst6, dst7);
         PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
         PCKEV_B2_UB(dst5, dst4, dst7, dst6, out2, out3);
-        ST6x4_UB(out0, out1, dst, dst_stride);
+        ST_W2(out0, 0, 2, dst, dst_stride);
+        ST_H2(out0, 2, 6, dst + 4, dst_stride);
+        ST_W2(out1, 0, 2, dst + 2 * dst_stride, dst_stride);
+        ST_H2(out1, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
         dst += (4 * dst_stride);
-        ST6x4_UB(out2, out3, dst, dst_stride);
+        ST_W2(out2, 0, 2, dst, dst_stride);
+        ST_H2(out2, 2, 6, dst + 4, dst_stride);
+        ST_W2(out3, 0, 2, dst + 2 * dst_stride, dst_stride);
+        ST_H2(out3, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -195,7 +201,7 @@  static void hevc_bi_copy_8w_msa(uint8_t *src0_ptr,
         SLLI_2V(dst0, dst1, 6);
         HEVC_BI_RND_CLIP2_MAX_SATU(in0, in1, dst0, dst1, 7, dst0, dst1);
         out0 = (v16u8) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-        ST8x2_UB(out0, dst, dst_stride);
+        ST_D2(out0, 0, 1, dst, dst_stride);
     } else if (4 == height) {
         LD4(src0_ptr, src_stride, tp0, tp1, tp2, tp3);
         INSERT_D2_SB(tp0, tp1, src0);
@@ -207,7 +213,7 @@  static void hevc_bi_copy_8w_msa(uint8_t *src0_ptr,
         HEVC_BI_RND_CLIP4_MAX_SATU(in0, in1, in2, in3, dst0, dst1, dst2, dst3,
                                    7, dst0, dst1, dst2, dst3);
         PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
     } else if (6 == height) {
         LD4(src0_ptr, src_stride, tp0, tp1, tp2, tp3);
         src0_ptr += 4 * src_stride;
@@ -225,9 +231,8 @@  static void hevc_bi_copy_8w_msa(uint8_t *src0_ptr,
                                    7, dst0, dst1, dst2, dst3);
         HEVC_BI_RND_CLIP2_MAX_SATU(in4, in5, dst4, dst5, 7, dst4, dst5);
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
-        ST8x4_UB(out0, out1, dst, dst_stride);
-        dst += (4 * dst_stride);
-        ST8x2_UB(out2, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+        ST_D2(out2, 0, 1, dst + 4 * dst_stride, dst_stride);
     } else if (0 == height % 8) {
         uint32_t loop_cnt;
 
@@ -255,7 +260,7 @@  static void hevc_bi_copy_8w_msa(uint8_t *src0_ptr,
                                        dst7, 7, dst4, dst5, dst6, dst7);
             PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
             PCKEV_B2_UB(dst5, dst4, dst7, dst6, out2, out3);
-            ST8x8_UB(out0, out1, out2, out3, dst, dst_stride);
+            ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride);
             dst += (8 * dst_stride);
         }
     }
@@ -294,7 +299,8 @@  static void hevc_bi_copy_12w_msa(uint8_t *src0_ptr,
                                    7, dst0, dst1, dst2, dst3);
         HEVC_BI_RND_CLIP2_MAX_SATU(in4, in5, dst4, dst5, 7, dst4, dst5);
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
-        ST12x4_UB(out0, out1, out2, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(out2, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -378,7 +384,7 @@  static void hevc_bi_copy_24w_msa(uint8_t *src0_ptr,
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
         PCKEV_B3_UB(dst7, dst6, dst9, dst8, dst11, dst10, out3, out4, out5);
         ST_UB4(out0, out1, out3, out4, dst, dst_stride);
-        ST8x4_UB(out2, out5, dst + 16, dst_stride);
+        ST_D4(out2, out5, 0, 1, 0, 1, dst + 16, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -588,7 +594,7 @@  static void hevc_hz_bi_8t_4w_msa(uint8_t *src0_ptr,
                           dst0, dst1, dst2, dst3, 7, dst0, dst1, dst2, dst3);
 
         PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
-        ST4x8_UB(dst0, dst1, dst, dst_stride);
+        ST_W8(dst0, dst1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
     }
 }
@@ -656,7 +662,7 @@  static void hevc_hz_bi_8t_8w_msa(uint8_t *src0_ptr,
                           dst0, dst1, dst2, dst3, 7, dst0, dst1, dst2, dst3);
 
         PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
-        ST8x4_UB(dst0, dst1, dst, dst_stride);
+        ST_D4(dst0, dst1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -1242,7 +1248,7 @@  static void hevc_vt_bi_8t_4w_msa(uint8_t *src0_ptr,
                           dst10, dst32, dst54, dst76);
 
         PCKEV_B2_SH(dst32, dst10, dst76, dst54, dst10, dst54);
-        ST4x8_UB(dst10, dst54, dst, dst_stride);
+        ST_W8(dst10, dst54, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
 
         src2110 = src10998;
@@ -1316,7 +1322,7 @@  static void hevc_vt_bi_8t_8w_msa(uint8_t *src0_ptr,
                           dst0_r, dst1_r, dst2_r, dst3_r);
 
         PCKEV_B2_SH(dst1_r, dst0_r, dst3_r, dst2_r, dst0_r, dst1_r);
-        ST8x4_UB(dst0_r, dst1_r, dst, dst_stride);
+        ST_D4(dst0_r, dst1_r, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         src10_r = src54_r;
@@ -1420,7 +1426,8 @@  static void hevc_vt_bi_8t_12w_msa(uint8_t *src0_ptr,
 
         PCKEV_B2_SH(dst1_r, dst0_r, dst3_r, dst2_r, dst0_r, dst1_r);
         dst0_l = (v8i16) __msa_pckev_b((v16i8) dst1_l, (v16i8) dst0_l);
-        ST12x4_UB(dst0_r, dst1_r, dst0_l, dst, dst_stride);
+        ST_D4(dst0_r, dst1_r, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(dst0_l, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
 
         src10_r = src54_r;
@@ -1721,7 +1728,7 @@  static void hevc_hv_bi_8t_4w_msa(uint8_t *src0_ptr,
         SRARI_H2_SH(out0, out1, 7);
         CLIP_SH2_0_255_MAX_SATU(out0, out1);
         out = (v16u8) __msa_pckev_b((v16i8) out1, (v16i8) out0);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         dst10 = dst54;
@@ -1849,7 +1856,7 @@  static void hevc_hv_bi_8t_8multx1mult_msa(uint8_t *src0_ptr,
             tmp = __msa_srari_h(tmp, 7);
             tmp = CLIP_SH_0_255_MAX_SATU(tmp);
             out = (v16u8) __msa_pckev_b((v16i8) tmp, (v16i8) tmp);
-            ST8x1_UB(out, dst_tmp);
+            ST_D1(out, 0, dst_tmp);
             dst_tmp += dst_stride;
 
             dst0 = dst1;
@@ -1995,7 +2002,7 @@  static void hevc_hv_bi_8t_12w_msa(uint8_t *src0_ptr,
         tmp = __msa_srari_h(tmp, 7);
         tmp = CLIP_SH_0_255_MAX_SATU(tmp);
         out = (v16u8) __msa_pckev_b((v16i8) tmp, (v16i8) tmp);
-        ST8x1_UB(out, dst_tmp);
+        ST_D1(out, 0, dst_tmp);
         dst_tmp += dst_stride;
 
         dst0 = dst1;
@@ -2083,7 +2090,7 @@  static void hevc_hv_bi_8t_12w_msa(uint8_t *src0_ptr,
         SRARI_H2_SH(out0, out1, 7);
         CLIP_SH2_0_255_MAX_SATU(out0, out1);
         out = (v16u8) __msa_pckev_b((v16i8) out1, (v16i8) out0);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         dst10 = dst54;
@@ -2211,7 +2218,7 @@  static void hevc_hz_bi_4t_4x2_msa(uint8_t *src0_ptr,
     tmp0 = CLIP_SH_0_255(tmp0);
     dst0 = __msa_pckev_b((v16i8) tmp0, (v16i8) tmp0);
 
-    ST4x2_UB(dst0, dst, dst_stride);
+    ST_W2(dst0, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hz_bi_4t_4x4_msa(uint8_t *src0_ptr,
@@ -2257,7 +2264,7 @@  static void hevc_hz_bi_4t_4x4_msa(uint8_t *src0_ptr,
     HEVC_BI_RND_CLIP2(in0, in1, tmp0, tmp1, 7, tmp0, tmp1);
     dst0 = __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
 
-    ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(dst0, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void hevc_hz_bi_4t_4x8multiple_msa(uint8_t *src0_ptr,
@@ -2318,7 +2325,7 @@  static void hevc_hz_bi_4t_4x8multiple_msa(uint8_t *src0_ptr,
                           tmp0, tmp1, tmp2, tmp3, 7, tmp0, tmp1, tmp2, tmp3);
 
         PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, dst0, dst1);
-        ST4x8_UB(dst0, dst1, dst, dst_stride);
+        ST_W8(dst0, dst1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
     }
 }
@@ -2398,7 +2405,10 @@  static void hevc_hz_bi_4t_6w_msa(uint8_t *src0_ptr,
                           dst0, dst1, dst2, dst3, 7, dst0, dst1, dst2, dst3);
 
         PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
-        ST6x4_UB(dst0, dst1, dst, dst_stride);
+        ST_W2(dst0, 0, 2, dst, dst_stride);
+        ST_H2(dst0, 2, 6, dst + 4, dst_stride);
+        ST_W2(dst1, 0, 2, dst + 2 * dst_stride, dst_stride);
+        ST_H2(dst1, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -2443,7 +2453,7 @@  static void hevc_hz_bi_4t_8x2_msa(uint8_t *src0_ptr,
     HEVC_BI_RND_CLIP2(in0, in1, dst0, dst1, 7, dst0, dst1);
 
     dst0 = (v8i16) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-    ST8x2_UB(dst0, dst, dst_stride);
+    ST_D2(dst0, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hz_bi_4t_8x6_msa(uint8_t *src0_ptr,
@@ -2506,9 +2516,8 @@  static void hevc_hz_bi_4t_8x6_msa(uint8_t *src0_ptr,
 
     PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
     dst2 = (v8i16) __msa_pckev_b((v16i8) dst5, (v16i8) dst4);
-    ST8x4_UB(dst0, dst1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST8x2_UB(dst2, dst, dst_stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, dst_stride);
+    ST_D2(dst2, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void hevc_hz_bi_4t_8x4multiple_msa(uint8_t *src0_ptr,
@@ -2564,7 +2573,7 @@  static void hevc_hz_bi_4t_8x4multiple_msa(uint8_t *src0_ptr,
                           dst0, dst1, dst2, dst3, 7, dst0, dst1, dst2, dst3);
 
         PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
-        ST8x4_UB(dst0, dst1, dst, dst_stride);
+        ST_D4(dst0, dst1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -2659,7 +2668,8 @@  static void hevc_hz_bi_4t_12w_msa(uint8_t *src0_ptr,
 
         PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
         dst2 = (v8i16) __msa_pckev_b((v16i8) dst5, (v16i8) dst4);
-        ST12x4_UB(dst0, dst1, dst2, dst, dst_stride);
+        ST_D4(dst0, dst1, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(dst2, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -2825,7 +2835,7 @@  static void hevc_hz_bi_4t_24w_msa(uint8_t *src0_ptr,
                           dst0, dst1, dst2, dst3, 7, dst0, dst1, dst2, dst3);
 
         PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
-        ST8x4_UB(dst0, dst1, dst_tmp, dst_stride);
+        ST_D4(dst0, dst1, 0, 1, 0, 1, dst_tmp, dst_stride);
         dst_tmp += (4 * dst_stride);
     }
 }
@@ -2936,7 +2946,7 @@  static void hevc_vt_bi_4t_4x2_msa(uint8_t *src0_ptr,
     dst10 = CLIP_SH_0_255(dst10);
 
     dst10 = (v8i16) __msa_pckev_b((v16i8) dst10, (v16i8) dst10);
-    ST4x2_UB(dst10, dst, dst_stride);
+    ST_W2(dst10, 0, 1, dst, dst_stride);
 }
 
 static void hevc_vt_bi_4t_4x4_msa(uint8_t *src0_ptr,
@@ -2985,7 +2995,7 @@  static void hevc_vt_bi_4t_4x4_msa(uint8_t *src0_ptr,
     HEVC_BI_RND_CLIP2(in0, in1, dst10, dst32, 7, dst10, dst32);
 
     dst10 = (v8i16) __msa_pckev_b((v16i8) dst32, (v16i8) dst10);
-    ST4x4_UB(dst10, dst10, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(dst10, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void hevc_vt_bi_4t_4x8multiple_msa(uint8_t *src0_ptr,
@@ -3056,7 +3066,7 @@  static void hevc_vt_bi_4t_4x8multiple_msa(uint8_t *src0_ptr,
                           dst10, dst32, dst54, dst76);
 
         PCKEV_B2_SH(dst32, dst10, dst76, dst54, dst10, dst54);
-        ST4x8_UB(dst10, dst54, dst, dst_stride);
+        ST_W8(dst10, dst54, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
     }
 }
@@ -3147,7 +3157,10 @@  static void hevc_vt_bi_4t_6w_msa(uint8_t *src0_ptr,
                       dst0_r, dst1_r, dst2_r, dst3_r);
 
     PCKEV_B2_SH(dst1_r, dst0_r, dst3_r, dst2_r, dst0_r, dst1_r);
-    ST6x4_UB(dst0_r, dst1_r, dst, dst_stride);
+    ST_W2(dst0_r, 0, 2, dst, dst_stride);
+    ST_H2(dst0_r, 2, 6, dst + 4, dst_stride);
+    ST_W2(dst1_r, 0, 2, dst + 2 * dst_stride, dst_stride);
+    ST_H2(dst1_r, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
     dst += (4 * dst_stride);
 
     LD_SH4(src1_ptr, src2_stride, in0, in1, in2, in3);
@@ -3171,7 +3184,10 @@  static void hevc_vt_bi_4t_6w_msa(uint8_t *src0_ptr,
                       dst0_r, dst1_r, dst2_r, dst3_r);
 
     PCKEV_B2_SH(dst1_r, dst0_r, dst3_r, dst2_r, dst0_r, dst1_r);
-    ST6x4_UB(dst0_r, dst1_r, dst, dst_stride);
+    ST_W2(dst0_r, 0, 2, dst, dst_stride);
+    ST_H2(dst0_r, 2, 6, dst + 4, dst_stride);
+    ST_W2(dst1_r, 0, 2, dst + 2 * dst_stride, dst_stride);
+    ST_H2(dst1_r, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
     dst += (4 * dst_stride);
 }
 
@@ -3216,7 +3232,7 @@  static void hevc_vt_bi_4t_8x2_msa(uint8_t *src0_ptr,
     HEVC_BI_RND_CLIP2(in0, in1, dst0_r, dst1_r, 7, dst0_r, dst1_r);
     dst0_r = (v8i16) __msa_pckev_b((v16i8) dst1_r, (v16i8) dst0_r);
 
-    ST8x2_UB(dst0_r, dst, dst_stride);
+    ST_D2(dst0_r, 0, 1, dst, dst_stride);
 }
 
 static void hevc_vt_bi_4t_8x6_msa(uint8_t *src0_ptr,
@@ -3275,9 +3291,8 @@  static void hevc_vt_bi_4t_8x6_msa(uint8_t *src0_ptr,
 
     PCKEV_B2_SH(dst1_r, dst0_r, dst3_r, dst2_r, dst0_r, dst1_r);
     dst2_r = (v8i16) __msa_pckev_b((v16i8) dst5_r, (v16i8) dst4_r);
-    ST8x4_UB(dst0_r, dst1_r, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST8x2_UB(dst2_r, dst, dst_stride);
+    ST_D4(dst0_r, dst1_r, 0, 1, 0, 1, dst, dst_stride);
+    ST_D2(dst2_r, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void hevc_vt_bi_4t_8x4multiple_msa(uint8_t *src0_ptr,
@@ -3337,7 +3352,7 @@  static void hevc_vt_bi_4t_8x4multiple_msa(uint8_t *src0_ptr,
                           dst0_r, dst1_r, dst2_r, dst3_r);
 
         PCKEV_B2_SH(dst1_r, dst0_r, dst3_r, dst2_r, dst0_r, dst1_r);
-        ST8x4_UB(dst0_r, dst1_r, dst, dst_stride);
+        ST_D4(dst0_r, dst1_r, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -3436,7 +3451,8 @@  static void hevc_vt_bi_4t_12w_msa(uint8_t *src0_ptr,
 
         PCKEV_B2_SH(dst1_r, dst0_r, dst3_r, dst2_r, dst0_r, dst1_r);
         dst0_l = (v8i16) __msa_pckev_b((v16i8) dst1_l, (v16i8) dst0_l);
-        ST12x4_UB(dst0_r, dst1_r, dst0_l, dst, dst_stride);
+        ST_D4(dst0_r, dst1_r, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(dst0_l, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
 
         src2 = src6;
@@ -3610,7 +3626,7 @@  static void hevc_vt_bi_4t_24w_msa(uint8_t *src0_ptr,
         PCKEV_B2_SH(dst0_l, dst0_r, dst1_l, dst1_r, dst0_r, dst1_r);
         dst2_r = (v8i16) __msa_pckev_b((v16i8) dst3_r, (v16i8) dst2_r);
         ST_SH2(dst0_r, dst1_r, dst, dst_stride);
-        ST8x2_UB(dst2_r, dst + 16, dst_stride);
+        ST_D2(dst2_r, 0, 1, dst + 16, dst_stride);
         dst += (2 * dst_stride);
 
         /* 16width */
@@ -3650,7 +3666,7 @@  static void hevc_vt_bi_4t_24w_msa(uint8_t *src0_ptr,
         PCKEV_B2_SH(dst0_l, dst0_r, dst1_l, dst1_r, dst0_r, dst1_r);
         dst2_r = (v8i16) __msa_pckev_b((v16i8) dst3_r, (v16i8) dst2_r);
         ST_SH2(dst0_r, dst1_r, dst, dst_stride);
-        ST8x2_UB(dst2_r, dst + 16, dst_stride);
+        ST_D2(dst2_r, 0, 1, dst + 16, dst_stride);
         dst += (2 * dst_stride);
     }
 }
@@ -3829,7 +3845,7 @@  static void hevc_hv_bi_4t_4x2_msa(uint8_t *src0_ptr,
     tmp = __msa_srari_h(tmp, 7);
     tmp = CLIP_SH_0_255_MAX_SATU(tmp);
     out = (v16u8) __msa_pckev_b((v16i8) tmp, (v16i8) tmp);
-    ST4x2_UB(out, dst, dst_stride);
+    ST_W2(out, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hv_bi_4t_4x4_msa(uint8_t *src0_ptr,
@@ -3905,7 +3921,7 @@  static void hevc_hv_bi_4t_4x4_msa(uint8_t *src0_ptr,
     SRARI_H2_SH(tmp0, tmp1, 7);
     CLIP_SH2_0_255_MAX_SATU(tmp0, tmp1);
     out = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void hevc_hv_bi_4t_4multx8mult_msa(uint8_t *src0_ptr,
@@ -4018,7 +4034,7 @@  static void hevc_hv_bi_4t_4multx8mult_msa(uint8_t *src0_ptr,
         SRARI_H4_SH(tmp0, tmp1, tmp2, tmp3, 7);
         CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
         PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST4x8_UB(out0, out1, dst, dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
 
         dst10_r = dst98_r;
@@ -4186,7 +4202,7 @@  static void hevc_hv_bi_4t_6w_msa(uint8_t *src0_ptr,
     SRARI_H4_SH(tmp0, tmp1, tmp2, tmp3, 7);
     CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
     PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-    ST4x8_UB(out0, out1, dst, dst_stride);
+    ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
 
     LW4(src1_ptr + 4, src2_stride, tpw0, tpw1, tpw2, tpw3);
     src1_ptr += (4 * src2_stride);
@@ -4198,9 +4214,7 @@  static void hevc_hv_bi_4t_6w_msa(uint8_t *src0_ptr,
     SRARI_H2_SH(tmp4, tmp5, 7);
     CLIP_SH2_0_255_MAX_SATU(tmp4, tmp5);
     out2 = (v16u8) __msa_pckev_b((v16i8) tmp5, (v16i8) tmp4);
-    ST2x4_UB(out2, 0, dst + 4, dst_stride);
-    dst += 4 * dst_stride;
-    ST2x4_UB(out2, 4, dst + 4, dst_stride);
+    ST_H8(out2, 0, 1, 2, 3, 4, 5, 6, 7, dst + 4, dst_stride);
 }
 
 static void hevc_hv_bi_4t_8x2_msa(uint8_t *src0_ptr,
@@ -4274,7 +4288,7 @@  static void hevc_hv_bi_4t_8x2_msa(uint8_t *src0_ptr,
     SRARI_H2_SH(tmp0, tmp1, 7);
     CLIP_SH2_0_255_MAX_SATU(tmp0, tmp1);
     out = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-    ST8x2_UB(out, dst, dst_stride);
+    ST_D2(out, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hv_bi_4t_8multx4_msa(uint8_t *src0_ptr,
@@ -4368,7 +4382,7 @@  static void hevc_hv_bi_4t_8multx4_msa(uint8_t *src0_ptr,
         SRARI_H4_SH(tmp0, tmp1, tmp2, tmp3, 7);
         CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
         PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
         dst += 8;
     }
 }
@@ -4485,9 +4499,8 @@  static void hevc_hv_bi_4t_8x6_msa(uint8_t *src0_ptr,
     CLIP_SH2_0_255_MAX_SATU(tmp4, tmp5);
     PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
     out2 = (v16u8) __msa_pckev_b((v16i8) tmp5, (v16i8) tmp4);
-    ST8x4_UB(out0, out1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST8x2_UB(out2, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+    ST_D2(out2, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void hevc_hv_bi_4t_8multx4mult_msa(uint8_t *src0_ptr,
@@ -4599,7 +4612,7 @@  static void hevc_hv_bi_4t_8multx4mult_msa(uint8_t *src0_ptr,
             SRARI_H4_SH(tmp0, tmp1, tmp2, tmp3, 7);
             CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
             PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-            ST8x4_UB(out0, out1, dst_tmp, dst_stride);
+            ST_D4(out0, out1, 0, 1, 0, 1, dst_tmp, dst_stride);
             dst_tmp += (4 * dst_stride);
 
             dst10_r = dst54_r;
@@ -4749,7 +4762,7 @@  static void hevc_hv_bi_4t_12w_msa(uint8_t *src0_ptr,
         SRARI_H4_SH(tmp0, tmp1, tmp2, tmp3, 7);
         CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
         PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST8x4_UB(out0, out1, dst_tmp, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst_tmp, dst_stride);
         dst_tmp += (4 * dst_stride);
 
         dst10_r = dst54_r;
@@ -4835,7 +4848,7 @@  static void hevc_hv_bi_4t_12w_msa(uint8_t *src0_ptr,
         SRARI_H4_SH(tmp0, tmp1, tmp2, tmp3, 7);
         CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
         PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST4x8_UB(out0, out1, dst, dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
 
         dst10_r = dst98_r;
diff --git a/libavcodec/mips/hevc_mc_biw_msa.c b/libavcodec/mips/hevc_mc_biw_msa.c
index ea65f00..68f122e 100644
--- a/libavcodec/mips/hevc_mc_biw_msa.c
+++ b/libavcodec/mips/hevc_mc_biw_msa.c
@@ -126,7 +126,7 @@  static void hevc_biwgt_copy_4w_msa(uint8_t *src0_ptr,
         dst0 = (v8i16) __msa_pckev_h((v8i16) dst0_l, (v8i16) dst0_r);
         dst0 = CLIP_SH_0_255_MAX_SATU(dst0);
         out0 = (v16u8) __msa_pckev_b((v16i8) dst0, (v16i8) dst0);
-        ST4x2_UB(out0, dst, dst_stride);
+        ST_W2(out0, 0, 1, dst, dst_stride);
     } else if (4 == height) {
         LW4(src0_ptr, src_stride, tp0, tp1, tp2, tp3);
         INSERT_W4_SB(tp0, tp1, tp2, tp3, src0);
@@ -138,7 +138,7 @@  static void hevc_biwgt_copy_4w_msa(uint8_t *src0_ptr,
         HEVC_BIW_RND_CLIP2_MAX_SATU(dst0, dst1, in0, in1, weight_vec, rnd_vec,
                                     offset_vec, dst0, dst1);
         out0 = (v16u8) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-        ST4x4_UB(out0, out0, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out0, 0, 1, 2, 3, dst, dst_stride);
     } else if (0 == height % 8) {
         for (loop_cnt = (height >> 3); loop_cnt--;) {
             LW4(src0_ptr, src_stride, tp0, tp1, tp2, tp3);
@@ -162,7 +162,7 @@  static void hevc_biwgt_copy_4w_msa(uint8_t *src0_ptr,
                                         in3, weight_vec, rnd_vec, offset_vec,
                                         dst0, dst1, dst2, dst3);
             PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-            ST4x8_UB(out0, out1, dst, dst_stride);
+            ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
             dst += (8 * dst_stride);
         }
     }
@@ -214,7 +214,10 @@  static void hevc_biwgt_copy_6w_msa(uint8_t *src0_ptr,
                                     weight_vec, rnd_vec, offset_vec,
                                     dst0, dst1, dst2, dst3);
         PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-        ST6x4_UB(out0, out1, dst, dst_stride);
+        ST_W2(out0, 0, 2, dst, dst_stride);
+        ST_H2(out0, 2, 6, dst + 4, dst_stride);
+        ST_W2(out1, 0, 2, dst + 2 * dst_stride, dst_stride);
+        ST_H2(out1, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -261,7 +264,7 @@  static void hevc_biwgt_copy_8w_msa(uint8_t *src0_ptr,
                            dst0, dst1);
 
         out0 = (v16u8) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-        ST8x2_UB(out0, dst, dst_stride);
+        ST_D2(out0, 0, 1, dst, dst_stride);
     } else if (6 == height) {
         LD4(src0_ptr, src_stride, tp0, tp1, tp2, tp3);
         src0_ptr += 4 * src_stride;
@@ -281,9 +284,8 @@  static void hevc_biwgt_copy_8w_msa(uint8_t *src0_ptr,
         HEVC_BIW_RND_CLIP2_MAX_SATU(dst4, dst5, in4, in5, weight_vec, rnd_vec,
                                     offset_vec, dst4, dst5);
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
-        ST8x4_UB(out0, out1, dst, dst_stride);
-        dst += (4 * dst_stride);
-        ST8x2_UB(out2, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+        ST_D2(out2, 0, 1, dst + 4 * dst_stride, dst_stride);
     } else if (0 == height % 4) {
         uint32_t loop_cnt;
 
@@ -302,7 +304,7 @@  static void hevc_biwgt_copy_8w_msa(uint8_t *src0_ptr,
                                         in3, weight_vec, rnd_vec, offset_vec,
                                         dst0, dst1, dst2, dst3);
             PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-            ST8x4_UB(out0, out1, dst, dst_stride);
+            ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
             dst += (4 * dst_stride);
         }
     }
@@ -361,7 +363,8 @@  static void hevc_biwgt_copy_12w_msa(uint8_t *src0_ptr,
         HEVC_BIW_RND_CLIP2_MAX_SATU(dst4, dst5, in4, in5, weight_vec, rnd_vec,
                                     offset_vec, dst4, dst5);
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
-        ST12x4_UB(out0, out1, out2, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(out2, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -480,7 +483,7 @@  static void hevc_biwgt_copy_24w_msa(uint8_t *src0_ptr,
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
         PCKEV_B3_UB(dst7, dst6, dst9, dst8, dst11, dst10, out3, out4, out5);
         ST_UB4(out0, out1, out3, out4, dst, dst_stride);
-        ST8x4_UB(out2, out5, dst + 16, dst_stride);
+        ST_D4(out2, out5, 0, 1, 0, 1, dst + 16, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -720,7 +723,7 @@  static void hevc_hz_biwgt_8t_4w_msa(uint8_t *src0_ptr,
                            out0, out1);
 
         out0 = (v8i16) __msa_pckev_b((v16i8) out1, (v16i8) out0);
-        ST4x4_UB(out0, out0, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out0, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -800,7 +803,7 @@  static void hevc_hz_biwgt_8t_8w_msa(uint8_t *src0_ptr,
                            out0, out1, out2, out3);
 
         PCKEV_B2_SH(out1, out0, out3, out2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -876,7 +879,7 @@  static void hevc_hz_biwgt_8t_12w_msa(uint8_t *src0_ptr,
                            weight_vec, rnd_vec, offset_vec, out0, out1, out2,
                            out3);
         PCKEV_B2_SH(out1, out0, out3, out2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 
         LD_SB4(src0_ptr + 8, src_stride, src0, src1, src2, src3);
         src0_ptr += (4 * src_stride);
@@ -895,7 +898,7 @@  static void hevc_hz_biwgt_8t_12w_msa(uint8_t *src0_ptr,
         HEVC_BIW_RND_CLIP2(dst0, dst1, in0, in1, weight_vec, rnd_vec,
                            offset_vec, out0, out1);
         out0 = (v8i16) __msa_pckev_b((v16i8) out1, (v16i8) out0);
-        ST4x4_UB(out0, out0, 0, 1, 2, 3, dst + 8, dst_stride);
+        ST_W4(out0, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -1483,7 +1486,7 @@  static void hevc_vt_biwgt_8t_4w_msa(uint8_t *src0_ptr,
                            out0, out1, out2, out3);
 
         PCKEV_B2_SH(out1, out0, out3, out2, out0, out1);
-        ST4x8_UB(out0, out1, dst, dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
 
         src2110 = src10998;
@@ -1568,7 +1571,7 @@  static void hevc_vt_biwgt_8t_8w_msa(uint8_t *src0_ptr,
                            out0, out1, out2, out3);
 
         PCKEV_B2_SH(out1, out0, out3, out2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         src10_r = src54_r;
@@ -1674,8 +1677,8 @@  static void hevc_vt_biwgt_8t_12w_msa(uint8_t *src0_ptr,
         dst2_r = (v4i32) __msa_pckev_h((v8i16) dst2_l, (v8i16) dst2_r);
         out2 = CLIP_SH_0_255(dst2_r);
         PCKEV_B2_SH(out1, out0, out2, out2, out0, out2);
-        ST8x2_UB(out0, dst, dst_stride);
-        ST4x2_UB(out2, dst + 8, dst_stride);
+        ST_D2(out0, 0, 1, dst, dst_stride);
+        ST_W2(out2, 0, 1, dst + 8, dst_stride);
         dst += (2 * dst_stride);
 
         src10_r = src32_r;
@@ -2048,7 +2051,7 @@  static void hevc_hv_biwgt_8t_4w_msa(uint8_t *src0_ptr,
         CLIP_SW4_0_255_MAX_SATU(dst0, dst1, dst2, dst3);
         PCKEV_H2_SH(dst1, dst0, dst3, dst2, tmp0, tmp1);
         out = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         dst10 = dst54;
@@ -2226,7 +2229,7 @@  static void hevc_hv_biwgt_8t_8multx2mult_msa(uint8_t *src0_ptr,
             CLIP_SW4_0_255_MAX_SATU(dst0_l, dst0_r, dst1_l, dst1_r);
             PCKEV_H2_SH(dst0_l, dst0_r, dst1_l, dst1_r, tmp0, tmp1);
             out = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-            ST8x2_UB(out, dst_tmp, dst_stride);
+            ST_D2(out, 0, 1, dst_tmp, dst_stride);
             dst_tmp += (2 * dst_stride);
 
             dst0 = dst2;
@@ -2412,7 +2415,7 @@  static void hevc_hv_biwgt_8t_12w_msa(uint8_t *src0_ptr,
         CLIP_SW4_0_255_MAX_SATU(dst1, dst0, dst3, dst2);
         PCKEV_H2_SH(dst1, dst0, dst3, dst2, tmp0, tmp1);
         out = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-        ST8x2_UB(out, dst_tmp, dst_stride);
+        ST_D2(out, 0, 1, dst_tmp, dst_stride);
         dst_tmp += (2 * dst_stride);
 
         dsth0 = dsth2;
@@ -2503,7 +2506,7 @@  static void hevc_hv_biwgt_8t_12w_msa(uint8_t *src0_ptr,
         CLIP_SW4_0_255_MAX_SATU(dst0, dst1, dst2, dst3);
         PCKEV_H2_SH(dst1, dst0, dst3, dst2, tmp0, tmp1);
         out = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         dst10 = dst54;
@@ -2683,7 +2686,7 @@  static void hevc_hz_biwgt_4t_4x2_msa(uint8_t *src0_ptr,
     dst0_r = (v4i32) __msa_pckev_h((v8i16) dst0_l, (v8i16) dst0_r);
     out0 = CLIP_SH_0_255(dst0_r);
     out0 = (v8i16) __msa_pckev_b((v16i8) out0, (v16i8) out0);
-    ST4x2_UB(out0, dst, dst_stride);
+    ST_W2(out0, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hz_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
@@ -2743,7 +2746,7 @@  static void hevc_hz_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
                        dst0, dst1);
 
     dst0 = (v8i16) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-    ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(dst0, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void hevc_hz_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr,
@@ -2816,7 +2819,7 @@  static void hevc_hz_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr,
                            dst0, dst1, dst2, dst3);
 
         PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
-        ST4x8_UB(dst0, dst1, dst, dst_stride);
+        ST_W8(dst0, dst1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
     }
 }
@@ -2918,7 +2921,10 @@  static void hevc_hz_biwgt_4t_6w_msa(uint8_t *src0_ptr,
                            dst0, dst1, dst2, dst3);
 
         PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
-        ST6x4_UB(dst0, dst1, dst, dst_stride);
+        ST_W2(dst0, 0, 2, dst, dst_stride);
+        ST_H2(dst0, 2, 6, dst + 4, dst_stride);
+        ST_W2(dst1, 0, 2, dst + 2 * dst_stride, dst_stride);
+        ST_H2(dst1, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -2976,7 +2982,7 @@  static void hevc_hz_biwgt_4t_8x2_msa(uint8_t *src0_ptr,
                        dst0, dst1);
 
     dst0 = (v8i16) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-    ST8x2_UB(dst0, dst, dst_stride);
+    ST_D2(dst0, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hz_biwgt_4t_8x6_msa(uint8_t *src0_ptr,
@@ -3049,9 +3055,8 @@  static void hevc_hz_biwgt_4t_8x6_msa(uint8_t *src0_ptr,
 
     PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
     dst3 = (v8i16) __msa_pckev_b((v16i8) dst5, (v16i8) dst4);
-    ST8x4_UB(dst0, dst1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST8x2_UB(dst3, dst, dst_stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, dst_stride);
+    ST_D2(dst3, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void hevc_hz_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr,
@@ -3119,7 +3124,7 @@  static void hevc_hz_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr,
                            dst0, dst1, dst2, dst3);
 
         PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
-        ST8x4_UB(dst0, dst1, dst, dst_stride);
+        ST_D4(dst0, dst1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -3235,7 +3240,8 @@  static void hevc_hz_biwgt_4t_12w_msa(uint8_t *src0_ptr,
 
         PCKEV_B2_SH(dst1, dst0, dst3, dst2, dst0, dst1);
         dst3 = (v8i16) __msa_pckev_b((v16i8) dst5, (v16i8) dst4);
-        ST12x4_UB(dst0, dst1, dst3, dst, dst_stride);
+        ST_D4(dst0, dst1, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(dst3, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -3411,7 +3417,7 @@  static void hevc_hz_biwgt_4t_24w_msa(uint8_t *src0_ptr,
                            dst0, dst1);
 
         dst0 = (v8i16) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-        ST8x2_UB(dst0, (dst + 16), dst_stride);
+        ST_D2(dst0, 0, 1, (dst + 16), dst_stride);
         dst += (2 * dst_stride);
     }
 }
@@ -3551,7 +3557,7 @@  static void hevc_vt_biwgt_4t_4x2_msa(uint8_t *src0_ptr,
     dst10_r = (v4i32) __msa_pckev_h((v8i16) dst10_l, (v8i16) dst10_r);
     out = CLIP_SH_0_255(dst10_r);
     out = (v8i16) __msa_pckev_b((v16i8) out, (v16i8) out);
-    ST4x2_UB(out, dst, dst_stride);
+    ST_W2(out, 0, 1, dst, dst_stride);
 }
 
 static void hevc_vt_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
@@ -3617,7 +3623,7 @@  static void hevc_vt_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
                        dst10, dst32);
 
     dst10 = (v8i16) __msa_pckev_b((v16i8) dst32, (v16i8) dst10);
-    ST4x4_UB(dst10, dst10, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(dst10, 0, 1, 2, 3, dst, dst_stride);
     dst += (4 * dst_stride);
 }
 
@@ -3702,7 +3708,7 @@  static void hevc_vt_biwgt_4t_4x8multiple_msa(uint8_t *src0_ptr,
                            dst10, dst32, dst54, dst76);
 
         PCKEV_B2_SH(dst32, dst10, dst76, dst54, dst10, dst32);
-        ST4x8_UB(dst10, dst32, dst, dst_stride);
+        ST_W8(dst10, dst32, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
     }
 }
@@ -3807,7 +3813,10 @@  static void hevc_vt_biwgt_4t_6w_msa(uint8_t *src0_ptr,
                            tmp0, tmp1, tmp2, tmp3);
 
         PCKEV_B2_SH(tmp1, tmp0, tmp3, tmp2, tmp0, tmp1);
-        ST6x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_W2(tmp0, 0, 2, dst, dst_stride);
+        ST_H2(tmp0, 2, 6, dst + 4, dst_stride);
+        ST_W2(tmp1, 0, 2, dst + 2 * dst_stride, dst_stride);
+        ST_H2(tmp1, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -3866,7 +3875,7 @@  static void hevc_vt_biwgt_4t_8x2_msa(uint8_t *src0_ptr,
                        tmp0, tmp1);
 
     tmp0 = (v8i16) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-    ST8x2_UB(tmp0, dst, dst_stride);
+    ST_D2(tmp0, 0, 1, dst, dst_stride);
 }
 
 static void hevc_vt_biwgt_4t_8x6_msa(uint8_t *src0_ptr,
@@ -3936,9 +3945,8 @@  static void hevc_vt_biwgt_4t_8x6_msa(uint8_t *src0_ptr,
 
     PCKEV_B2_SH(tmp1, tmp0, tmp3, tmp2, tmp0, tmp1);
     tmp3 = (v8i16) __msa_pckev_b((v16i8) tmp5, (v16i8) tmp4);
-    ST8x4_UB(tmp0, tmp1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST8x2_UB(tmp3, dst, dst_stride);
+    ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
+    ST_D2(tmp3, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void hevc_vt_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr,
@@ -4010,7 +4018,7 @@  static void hevc_vt_biwgt_4t_8x4multiple_msa(uint8_t *src0_ptr,
                            tmp0, tmp1, tmp2, tmp3);
 
         PCKEV_B2_SH(tmp1, tmp0, tmp3, tmp2, tmp0, tmp1);
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -4132,7 +4140,8 @@  static void hevc_vt_biwgt_4t_12w_msa(uint8_t *src0_ptr,
 
         PCKEV_B2_SH(tmp1, tmp0, tmp3, tmp2, tmp0, tmp1);
         tmp2 = (v8i16) __msa_pckev_b((v16i8) tmp5, (v16i8) tmp4);
-        ST12x4_UB(tmp0, tmp1, tmp2, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(tmp2, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -4323,7 +4332,7 @@  static void hevc_vt_biwgt_4t_24w_msa(uint8_t *src0_ptr,
         /* 8width */
         tmp2 = (v8i16) __msa_pckev_b((v16i8) tmp3, (v16i8) tmp2);
         ST_SH2(tmp0, tmp1, dst, dst_stride);
-        ST8x2_UB(tmp2, dst + 16, dst_stride);
+        ST_D2(tmp2, 0, 1, dst + 16, dst_stride);
         dst += (2 * dst_stride);
 
         /* 16width */
@@ -4363,7 +4372,7 @@  static void hevc_vt_biwgt_4t_24w_msa(uint8_t *src0_ptr,
         /* 8width */
         tmp2 = (v8i16) __msa_pckev_b((v16i8) tmp3, (v16i8) tmp2);
         ST_SH2(tmp0, tmp1, dst, dst_stride);
-        ST8x2_UB(tmp2, dst + 16, dst_stride);
+        ST_D2(tmp2, 0, 1, dst + 16, dst_stride);
         dst += (2 * dst_stride);
     }
 }
@@ -4568,7 +4577,7 @@  static void hevc_hv_biwgt_4t_4x2_msa(uint8_t *src0_ptr,
     tmp = __msa_pckev_h((v8i16) dst1, (v8i16) dst0);
     tmp = CLIP_SH_0_255_MAX_SATU(tmp);
     out = (v16u8) __msa_pckev_b((v16i8) tmp, (v16i8) tmp);
-    ST4x2_UB(out, dst, dst_stride);
+    ST_W2(out, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hv_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
@@ -4665,7 +4674,7 @@  static void hevc_hv_biwgt_4t_4x4_msa(uint8_t *src0_ptr,
     PCKEV_H2_SH(dst1, dst0, dst3, dst2, tmp0, tmp1);
     CLIP_SH2_0_255_MAX_SATU(tmp0, tmp1);
     out = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void hevc_hv_biwgt_4t_4multx8mult_msa(uint8_t *src0_ptr,
@@ -4803,7 +4812,7 @@  static void hevc_hv_biwgt_4t_4multx8mult_msa(uint8_t *src0_ptr,
                     tmp2, tmp3);
         CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
         PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST4x8_UB(out0, out1, dst, dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
 
         dst10_r = dst98_r;
@@ -5001,7 +5010,7 @@  static void hevc_hv_biwgt_4t_6w_msa(uint8_t *src0_ptr,
                 tmp2, tmp3);
     CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
     PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-    ST4x8_UB(out0, out1, dst, dst_stride);
+    ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
 
     PCKEV_H2_SW(dst1_l, dst0_l, dst3_l, dst2_l, dst4, dst5);
 
@@ -5023,9 +5032,7 @@  static void hevc_hv_biwgt_4t_6w_msa(uint8_t *src0_ptr,
 
     CLIP_SH2_0_255_MAX_SATU(tmp4, tmp5);
     out2 = (v16u8) __msa_pckev_b((v16i8) tmp5, (v16i8) tmp4);
-    ST2x4_UB(out2, 0, dst + 4, dst_stride);
-    dst += 4 * dst_stride;
-    ST2x4_UB(out2, 4, dst + 4, dst_stride);
+    ST_H8(out2, 0, 1, 2, 3, 4, 5, 6, 7, dst + 4, dst_stride);
 }
 
 static void hevc_hv_biwgt_4t_8x2_msa(uint8_t *src0_ptr,
@@ -5121,7 +5128,7 @@  static void hevc_hv_biwgt_4t_8x2_msa(uint8_t *src0_ptr,
     PCKEV_H2_SH(dst0_l, dst0_r, dst1_l, dst1_r, tmp0, tmp1);
     CLIP_SH2_0_255_MAX_SATU(tmp0, tmp1);
     out = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-    ST8x2_UB(out, dst, dst_stride);
+    ST_D2(out, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hv_biwgt_4t_8multx4_msa(uint8_t *src0_ptr,
@@ -5243,7 +5250,7 @@  static void hevc_hv_biwgt_4t_8multx4_msa(uint8_t *src0_ptr,
                     tmp0, tmp1, tmp2, tmp3);
         CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
         PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
         dst += 8;
     }
 }
@@ -5394,9 +5401,8 @@  static void hevc_hv_biwgt_4t_8x6_msa(uint8_t *src0_ptr,
     PCKEV_H2_SH(dst1, dst0, dst3, dst2, tmp4, tmp5);
     CLIP_SH2_0_255_MAX_SATU(tmp4, tmp5);
     out2 = (v16u8) __msa_pckev_b((v16i8) tmp5, (v16i8) tmp4);
-    ST8x4_UB(out0, out1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST8x2_UB(out2, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+    ST_D2(out2, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void hevc_hv_biwgt_4t_8multx4mult_msa(uint8_t *src0_ptr,
@@ -5533,7 +5539,7 @@  static void hevc_hv_biwgt_4t_8multx4mult_msa(uint8_t *src0_ptr,
                         tmp0, tmp1, tmp2, tmp3);
             CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
             PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-            ST8x4_UB(out0, out1, dst_tmp, dst_stride);
+            ST_D4(out0, out1, 0, 1, 0, 1, dst_tmp, dst_stride);
             dst_tmp += (4 * dst_stride);
 
             dst10_r = dst54_r;
@@ -5720,7 +5726,7 @@  static void hevc_hv_biwgt_4t_12w_msa(uint8_t *src0_ptr,
                     tmp0, tmp1, tmp2, tmp3);
         CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
         PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST8x4_UB(out0, out1, dst_tmp, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst_tmp, dst_stride);
         dst_tmp += (4 * dst_stride);
 
         dst10_r = dst54_r;
@@ -5816,7 +5822,7 @@  static void hevc_hv_biwgt_4t_12w_msa(uint8_t *src0_ptr,
                     tmp0, tmp1, tmp2, tmp3);
         CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
         PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST4x8_UB(out0, out1, dst, dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
 
         dst10_r = dst98_r;
diff --git a/libavcodec/mips/hevc_mc_uni_msa.c b/libavcodec/mips/hevc_mc_uni_msa.c
index 740c970..36e6552 100644
--- a/libavcodec/mips/hevc_mc_uni_msa.c
+++ b/libavcodec/mips/hevc_mc_uni_msa.c
@@ -309,7 +309,7 @@  static void common_hz_8t_4x4_msa(uint8_t *src, int32_t src_stride,
     SRARI_H2_SH(out0, out1, 6);
     SAT_SH2_SH(out0, out1, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void common_hz_8t_4x8_msa(uint8_t *src, int32_t src_stride,
@@ -344,10 +344,9 @@  static void common_hz_8t_4x8_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_SH(out0, out1, out2, out3, 6);
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
     out = PCKEV_XORI128_UB(out2, out3);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
 }
 
 static void common_hz_8t_4x16_msa(uint8_t *src, int32_t src_stride,
@@ -382,11 +381,10 @@  static void common_hz_8t_4x16_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_SH(out0, out1, out2, out3, 6);
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
     out = PCKEV_XORI128_UB(out2, out3);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
+    dst += (8 * dst_stride);
 
     LD_SB4(src, src_stride, src0, src1, src2, src3);
     XORI_B4_128_SB(src0, src1, src2, src3);
@@ -402,10 +400,9 @@  static void common_hz_8t_4x16_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_SH(out0, out1, out2, out3, 6);
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
     out = PCKEV_XORI128_UB(out2, out3);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
 }
 
 static void common_hz_8t_4w_msa(uint8_t *src, int32_t src_stride,
@@ -468,7 +465,7 @@  static void common_hz_8t_8w_msa(uint8_t *src, int32_t src_stride,
         SAT_SH4_SH(out0, out1, out2, out3, 7);
         tmp0 = PCKEV_XORI128_UB(out0, out1);
         tmp1 = PCKEV_XORI128_UB(out2, out3);
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -546,8 +543,8 @@  static void common_hz_8t_12w_msa(uint8_t *src, int32_t src_stride,
         tmp1 = PCKEV_XORI128_UB(out2, out3);
         tmp2 = PCKEV_XORI128_UB(out4, out5);
 
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
-        ST4x4_UB(tmp2, tmp2, 0, 1, 2, 3, dst + 8, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(tmp2, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -670,7 +667,7 @@  static void common_hz_8t_24w_msa(uint8_t *src, int32_t src_stride,
         SAT_SH4_SH(out0, out8, out2, out9, 7);
         SAT_SH2_SH(out1, out3, 7);
         out = PCKEV_XORI128_UB(out8, out9);
-        ST8x2_UB(out, dst + 16, dst_stride);
+        ST_D2(out, 0, 1, dst + 16, dst_stride);
         out = PCKEV_XORI128_UB(out0, out1);
         ST_UB(out, dst);
         dst += dst_stride;
@@ -965,10 +962,8 @@  static void common_vt_8t_4w_msa(uint8_t *src, int32_t src_stride,
         SAT_SH2_SH(out54, out76, 7);
         out0 = PCKEV_XORI128_UB(out10, out32);
         out1 = PCKEV_XORI128_UB(out54, out76);
-        ST4x4_UB(out0, out0, 0, 1, 2, 3, dst, dst_stride);
-        dst += (4 * dst_stride);
-        ST4x4_UB(out1, out1, 0, 1, 2, 3, dst, dst_stride);
-        dst += (4 * dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
+        dst += (8 * dst_stride);
 
         src2110 = src10998;
         src4332 = src12111110;
@@ -1019,7 +1014,7 @@  static void common_vt_8t_8w_msa(uint8_t *src, int32_t src_stride,
         SAT_SH4_SH(out0_r, out1_r, out2_r, out3_r, 7);
         tmp0 = PCKEV_XORI128_UB(out0_r, out1_r);
         tmp1 = PCKEV_XORI128_UB(out2_r, out3_r);
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         src10_r = src54_r;
@@ -1458,10 +1453,8 @@  static void hevc_hv_uni_8t_4w_msa(uint8_t *src,
         PCKEV_H2_SW(dst5_r, dst4_r, dst7_r, dst6_r, dst4_r, dst5_r);
         out0 = PCKEV_XORI128_UB(dst0_r, dst1_r);
         out1 = PCKEV_XORI128_UB(dst4_r, dst5_r);
-        ST4x4_UB(out0, out0, 0, 1, 2, 3, dst, dst_stride);
-        dst += (4 * dst_stride);
-        ST4x4_UB(out1, out1, 0, 1, 2, 3, dst, dst_stride);
-        dst += (4 * dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
+        dst += (8 * dst_stride);
 
         dst10_r = dst98_r;
         dst32_r = dst1110_r;
@@ -1595,7 +1588,7 @@  static void hevc_hv_uni_8t_8multx2mult_msa(uint8_t *src,
 
             PCKEV_H2_SH(dst0_l, dst0_r, dst1_l, dst1_r, dst0, dst1);
             out = PCKEV_XORI128_UB(dst0, dst1);
-            ST8x2_UB(out, dst_tmp, dst_stride);
+            ST_D2(out, 0, 1, dst_tmp, dst_stride);
             dst_tmp += (2 * dst_stride);
 
             dst0 = dst2;
@@ -1741,7 +1734,7 @@  static void hevc_hv_uni_8t_12w_msa(uint8_t *src,
 
         PCKEV_H2_SH(dst0_l, dst0_r, dst1_l, dst1_r, dst0, dst1);
         out0 = PCKEV_XORI128_UB(dst0, dst1);
-        ST8x2_UB(out0, dst_tmp, dst_stride);
+        ST_D2(out0, 0, 1, dst_tmp, dst_stride);
         dst_tmp += (2 * dst_stride);
 
         dst0 = dst2;
@@ -1845,10 +1838,8 @@  static void hevc_hv_uni_8t_12w_msa(uint8_t *src,
         PCKEV_H2_SW(dst5_r, dst4_r, dst7_r, dst6_r, dst4_r, dst5_r);
         out0 = PCKEV_XORI128_UB(dst0_r, dst1_r);
         out1 = PCKEV_XORI128_UB(dst4_r, dst5_r);
-        ST4x4_UB(out0, out0, 0, 1, 2, 3, dst, dst_stride);
-        dst += (4 * dst_stride);
-        ST4x4_UB(out1, out1, 0, 1, 2, 3, dst, dst_stride);
-        dst += (4 * dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
+        dst += (8 * dst_stride);
 
         dst10_r = dst98_r;
         dst32_r = dst1110_r;
@@ -1944,7 +1935,7 @@  static void common_hz_4t_4x2_msa(uint8_t *src, int32_t src_stride,
     res0 = __msa_srari_h(res0, 6);
     res0 = __msa_sat_s_h(res0, 7);
     out = PCKEV_XORI128_UB(res0, res0);
-    ST4x2_UB(out, dst, dst_stride);
+    ST_W2(out, 0, 1, dst, dst_stride);
 }
 
 static void common_hz_4t_4x4_msa(uint8_t *src, int32_t src_stride,
@@ -1971,7 +1962,7 @@  static void common_hz_4t_4x4_msa(uint8_t *src, int32_t src_stride,
     SRARI_H2_SH(out0, out1, 6);
     SAT_SH2_SH(out0, out1, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void common_hz_4t_4x8_msa(uint8_t *src, int32_t src_stride,
@@ -2004,10 +1995,9 @@  static void common_hz_4t_4x8_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_SH(out0, out1, out2, out3, 6);
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
     out = PCKEV_XORI128_UB(out2, out3);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
 }
 
 static void common_hz_4t_4x16_msa(uint8_t *src, int32_t src_stride,
@@ -2038,11 +2028,10 @@  static void common_hz_4t_4x16_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_SH(out0, out1, out2, out3, 6);
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
     out = PCKEV_XORI128_UB(out2, out3);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
+    dst += (8 * dst_stride);
 
     LD_SB8(src, src_stride, src0, src1, src2, src3, src4, src5, src6, src7);
     src += (8 * src_stride);
@@ -2054,10 +2043,9 @@  static void common_hz_4t_4x16_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_SH(out0, out1, out2, out3, 6);
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
     out = PCKEV_XORI128_UB(out2, out3);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
 }
 
 static void common_hz_4t_4w_msa(uint8_t *src, int32_t src_stride,
@@ -2102,7 +2090,10 @@  static void common_hz_4t_6w_msa(uint8_t *src, int32_t src_stride,
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     out4 = PCKEV_XORI128_UB(out0, out1);
     out5 = PCKEV_XORI128_UB(out2, out3);
-    ST6x4_UB(out4, out5, dst, dst_stride);
+    ST_W2(out4, 0, 2, dst, dst_stride);
+    ST_H2(out4, 2, 6, dst + 4, dst_stride);
+    ST_W2(out5, 0, 2, dst + 2 * dst_stride, dst_stride);
+    ST_H2(out5, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
     dst += (4 * dst_stride);
 
     LD_SB4(src, src_stride, src0, src1, src2, src3);
@@ -2115,8 +2106,10 @@  static void common_hz_4t_6w_msa(uint8_t *src, int32_t src_stride,
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     out4 = PCKEV_XORI128_UB(out0, out1);
     out5 = PCKEV_XORI128_UB(out2, out3);
-    ST6x4_UB(out4, out5, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_W2(out4, 0, 2, dst, dst_stride);
+    ST_H2(out4, 2, 6, dst + 4, dst_stride);
+    ST_W2(out5, 0, 2, dst + 2 * dst_stride, dst_stride);
+    ST_H2(out5, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
 }
 
 static void common_hz_4t_8x2mult_msa(uint8_t *src, int32_t src_stride,
@@ -2148,7 +2141,7 @@  static void common_hz_4t_8x2mult_msa(uint8_t *src, int32_t src_stride,
         SRARI_H2_SH(vec0, vec1, 6);
         SAT_SH2_SH(vec0, vec1, 7);
         out = PCKEV_XORI128_UB(vec0, vec1);
-        ST8x2_UB(out, dst, dst_stride);
+        ST_D2(out, 0, 1, dst, dst_stride);
         dst += (2 * dst_stride);
     }
 }
@@ -2182,7 +2175,7 @@  static void common_hz_4t_8x4mult_msa(uint8_t *src, int32_t src_stride,
         SAT_SH4_SH(out0, out1, out2, out3, 7);
         tmp0 = PCKEV_XORI128_UB(out0, out1);
         tmp1 = PCKEV_XORI128_UB(out2, out3);
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -2235,7 +2228,7 @@  static void common_hz_4t_12w_msa(uint8_t *src, int32_t src_stride,
         SRARI_H2_SH(out0, out1, 6);
         SAT_SH2_SH(out0, out1, 7);
         tmp0 = PCKEV_XORI128_UB(out0, out1);
-        ST4x4_UB(tmp0, tmp0, 0, 1, 2, 3, dst + 8, dst_stride);
+        ST_W4(tmp0, 0, 1, 2, 3, dst + 8, dst_stride);
 
         VSHF_B2_SB(src0, src0, src1, src1, mask0, mask0, vec4, vec5);
         VSHF_B2_SB(src2, src2, src3, src3, mask0, mask0, vec6, vec7);
@@ -2249,7 +2242,7 @@  static void common_hz_4t_12w_msa(uint8_t *src, int32_t src_stride,
         SAT_SH4_SH(out2, out3, out4, out5, 7);
         tmp0 = PCKEV_XORI128_UB(out2, out3);
         tmp1 = PCKEV_XORI128_UB(out4, out5);
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -2395,7 +2388,7 @@  static void common_hz_4t_24w_msa(uint8_t *src, int32_t src_stride,
         SAT_SH4_SH(out0, out1, out2, out3, 7);
         tmp0 = PCKEV_XORI128_UB(out0, out1);
         tmp1 = PCKEV_XORI128_UB(out2, out3);
-        ST8x4_UB(tmp0, tmp1, dst1, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst1, dst_stride);
         dst1 += (4 * dst_stride);
     }
 }
@@ -2496,7 +2489,7 @@  static void common_vt_4t_4x2_msa(uint8_t *src, int32_t src_stride,
     out10 = __msa_srari_h(out10, 6);
     out10 = __msa_sat_s_h(out10, 7);
     out = PCKEV_XORI128_UB(out10, out10);
-    ST4x2_UB(out, dst, dst_stride);
+    ST_W2(out, 0, 1, dst, dst_stride);
 }
 
 static void common_vt_4t_4x4multiple_msa(uint8_t *src, int32_t src_stride,
@@ -2540,7 +2533,7 @@  static void common_vt_4t_4x4multiple_msa(uint8_t *src, int32_t src_stride,
         SRARI_H2_SH(out10, out32, 6);
         SAT_SH2_SH(out10, out32, 7);
         out = PCKEV_XORI128_UB(out10, out32);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -2596,7 +2589,10 @@  static void common_vt_4t_6w_msa(uint8_t *src, int32_t src_stride,
     SAT_SH4_SH(dst0_r, dst1_r, dst2_r, dst3_r, 7);
     out0 = PCKEV_XORI128_UB(dst0_r, dst1_r);
     out1 = PCKEV_XORI128_UB(dst2_r, dst3_r);
-    ST6x4_UB(out0, out1, dst, dst_stride);
+    ST_W2(out0, 0, 2, dst, dst_stride);
+    ST_H2(out0, 2, 6, dst + 4, dst_stride);
+    ST_W2(out1, 0, 2, dst + 2 * dst_stride, dst_stride);
+    ST_H2(out1, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
     dst += (4 * dst_stride);
 
     LD_SB2(src, src_stride, src3, src4);
@@ -2619,7 +2615,10 @@  static void common_vt_4t_6w_msa(uint8_t *src, int32_t src_stride,
     SAT_SH4_SH(dst0_r, dst1_r, dst2_r, dst3_r, 7);
     out0 = PCKEV_XORI128_UB(dst0_r, dst1_r);
     out1 = PCKEV_XORI128_UB(dst2_r, dst3_r);
-    ST6x4_UB(out0, out1, dst, dst_stride);
+    ST_W2(out0, 0, 2, dst, dst_stride);
+    ST_H2(out0, 2, 6, dst + 4, dst_stride);
+    ST_W2(out1, 0, 2, dst + 2 * dst_stride, dst_stride);
+    ST_H2(out1, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
 }
 
 static void common_vt_4t_8x2_msa(uint8_t *src, int32_t src_stride,
@@ -2645,7 +2644,7 @@  static void common_vt_4t_8x2_msa(uint8_t *src, int32_t src_stride,
     SRARI_H2_SH(tmp0, tmp1, 6);
     SAT_SH2_SH(tmp0, tmp1, 7);
     out = PCKEV_XORI128_UB(tmp0, tmp1);
-    ST8x2_UB(out, dst, dst_stride);
+    ST_D2(out, 0, 1, dst, dst_stride);
 }
 
 static void common_vt_4t_8x6_msa(uint8_t *src, int32_t src_stride,
@@ -2737,7 +2736,7 @@  static void common_vt_4t_8x4mult_msa(uint8_t *src, int32_t src_stride,
         SAT_SH4_SH(out0_r, out1_r, out2_r, out3_r, 7);
         tmp0 = PCKEV_XORI128_UB(out0_r, out1_r);
         tmp1 = PCKEV_XORI128_UB(out2_r, out3_r);
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         src10_r = src98_r;
@@ -2811,9 +2810,9 @@  static void common_vt_4t_12w_msa(uint8_t *src, int32_t src_stride,
         SAT_SH2_SH(dst0_l, dst1_l, 7);
         out0 = PCKEV_XORI128_UB(dst0_r, dst1_r);
         out1 = PCKEV_XORI128_UB(dst2_r, dst3_r);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
         out0 = PCKEV_XORI128_UB(dst0_l, dst1_l);
-        ST4x4_UB(out0, out0, 0, 1, 2, 3, dst + 8, dst_stride);
+        ST_W4(out0, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
 
         src2 = src6;
@@ -2982,12 +2981,12 @@  static void common_vt_4t_24w_msa(uint8_t *src, int32_t src_stride,
         out = PCKEV_XORI128_UB(out0_r, out0_l);
         ST_UB(out, dst);
         out = PCKEV_XORI128_UB(out2_r, out2_r);
-        ST8x1_UB(out, dst + 16);
+        ST_D1(out, 0, dst + 16);
         dst += dst_stride;
         out = PCKEV_XORI128_UB(out1_r, out1_l);
         ST_UB(out, dst);
         out = PCKEV_XORI128_UB(out3_r, out3_r);
-        ST8x1_UB(out, dst + 16);
+        ST_D1(out, 0, dst + 16);
         dst += dst_stride;
     }
 }
@@ -3137,7 +3136,7 @@  static void hevc_hv_uni_4t_4x2_msa(uint8_t *src,
     tmp = __msa_srari_h(tmp, 6);
     tmp = __msa_sat_s_h(tmp, 7);
     out = PCKEV_XORI128_UB(tmp, tmp);
-    ST4x2_UB(out, dst, dst_stride);
+    ST_W2(out, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hv_uni_4t_4x4_msa(uint8_t *src,
@@ -3196,7 +3195,7 @@  static void hevc_hv_uni_4t_4x4_msa(uint8_t *src,
     SRARI_H2_SH(tmp0, tmp1, 6);
     SAT_SH2_SH(tmp0, tmp1, 7);
     out = PCKEV_XORI128_UB(tmp0, tmp1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void hevc_hv_uni_4t_4multx8mult_msa(uint8_t *src,
@@ -3288,7 +3287,7 @@  static void hevc_hv_uni_4t_4multx8mult_msa(uint8_t *src,
         SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
         out0 = PCKEV_XORI128_UB(tmp0, tmp1);
         out1 = PCKEV_XORI128_UB(tmp2, tmp3);
-        ST4x8_UB(out0, out1, dst, dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
 
         dst10_r = dst98_r;
@@ -3432,10 +3431,8 @@  static void hevc_hv_uni_4t_6w_msa(uint8_t *src,
     out0 = PCKEV_XORI128_UB(tmp0, tmp1);
     out1 = PCKEV_XORI128_UB(tmp2, tmp3);
     out2 = PCKEV_XORI128_UB(tmp4, tmp5);
-    ST4x8_UB(out0, out1, dst, dst_stride);
-    ST2x4_UB(out2, 0, dst + 4, dst_stride);
-    dst += 4 * dst_stride;
-    ST2x4_UB(out2, 4, dst + 4, dst_stride);
+    ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
+    ST_H8(out2, 0, 1, 2, 3, 4, 5, 6, 7, dst + 4, dst_stride);
 }
 
 static void hevc_hv_uni_4t_8x2_msa(uint8_t *src,
@@ -3497,7 +3494,7 @@  static void hevc_hv_uni_4t_8x2_msa(uint8_t *src,
     SRARI_H2_SH(out0_r, out1_r, 6);
     SAT_SH2_SH(out0_r, out1_r, 7);
     out = PCKEV_XORI128_UB(out0_r, out1_r);
-    ST8x2_UB(out, dst, dst_stride);
+    ST_D2(out, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hv_uni_4t_8multx4_msa(uint8_t *src,
@@ -3580,7 +3577,7 @@  static void hevc_hv_uni_4t_8multx4_msa(uint8_t *src,
         SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
         out0 = PCKEV_XORI128_UB(tmp0, tmp1);
         out1 = PCKEV_XORI128_UB(tmp2, tmp3);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
         dst += 8;
     }
 }
@@ -3684,9 +3681,8 @@  static void hevc_hv_uni_4t_8x6_msa(uint8_t *src,
     out1 = PCKEV_XORI128_UB(out2_r, out3_r);
     out2 = PCKEV_XORI128_UB(out4_r, out5_r);
 
-    ST8x4_UB(out0, out1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST8x2_UB(out2, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+    ST_D2(out2, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void hevc_hv_uni_4t_8multx4mult_msa(uint8_t *src,
@@ -3788,7 +3784,7 @@  static void hevc_hv_uni_4t_8multx4mult_msa(uint8_t *src,
             SAT_SH4_SH(out0_r, out1_r, out2_r, out3_r, 7);
             out0 = PCKEV_XORI128_UB(out0_r, out1_r);
             out1 = PCKEV_XORI128_UB(out2_r, out3_r);
-            ST8x4_UB(out0, out1, dst_tmp, dst_stride);
+            ST_D4(out0, out1, 0, 1, 0, 1, dst_tmp, dst_stride);
             dst_tmp += (4 * dst_stride);
 
             dst10_r = dst54_r;
@@ -3919,7 +3915,7 @@  static void hevc_hv_uni_4t_12w_msa(uint8_t *src,
         SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
         out0 = PCKEV_XORI128_UB(tmp0, tmp1);
         out1 = PCKEV_XORI128_UB(tmp2, tmp3);
-        ST8x4_UB(out0, out1, dst_tmp, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst_tmp, dst_stride);
         dst_tmp += (4 * dst_stride);
 
         dst10_r = dst54_r;
@@ -3985,7 +3981,7 @@  static void hevc_hv_uni_4t_12w_msa(uint8_t *src,
         SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
         out0 = PCKEV_XORI128_UB(tmp0, tmp1);
         out1 = PCKEV_XORI128_UB(tmp2, tmp3);
-        ST4x8_UB(out0, out1, dst, dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
 
         dst10_r = dst98_r;
diff --git a/libavcodec/mips/hevc_mc_uniw_msa.c b/libavcodec/mips/hevc_mc_uniw_msa.c
index f9ecb41..cad1240 100644
--- a/libavcodec/mips/hevc_mc_uniw_msa.c
+++ b/libavcodec/mips/hevc_mc_uniw_msa.c
@@ -90,7 +90,7 @@  static void hevc_uniwgt_copy_4w_msa(uint8_t *src,
         dst0 += offset_vec;
         dst0 = CLIP_SH_0_255_MAX_SATU(dst0);
         out0 = (v16u8) __msa_pckev_b((v16i8) dst0, (v16i8) dst0);
-        ST4x2_UB(out0, dst, dst_stride);
+        ST_W2(out0, 0, 1, dst, dst_stride);
     } else if (4 == height) {
         LW4(src, src_stride, tp0, tp1, tp2, tp3);
         INSERT_W4_SB(tp0, tp1, tp2, tp3, src0);
@@ -99,7 +99,7 @@  static void hevc_uniwgt_copy_4w_msa(uint8_t *src,
         HEVC_UNIW_RND_CLIP2_MAX_SATU_H(dst0, dst1, weight_vec, offset_vec,
                                        rnd_vec, dst0, dst1);
         out0 = (v16u8) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-        ST4x4_UB(out0, out0, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out0, 0, 1, 2, 3, dst, dst_stride);
     } else if (0 == (height % 8)) {
         for (loop_cnt = (height >> 3); loop_cnt--;) {
             LW4(src, src_stride, tp0, tp1, tp2, tp3);
@@ -115,7 +115,7 @@  static void hevc_uniwgt_copy_4w_msa(uint8_t *src,
                                            offset_vec, rnd_vec, dst0, dst1,
                                            dst2, dst3);
             PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-            ST4x8_UB(out0, out1, dst, dst_stride);
+            ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
             dst += 8 * dst_stride;
         }
     }
@@ -170,9 +170,15 @@  static void hevc_uniwgt_copy_6w_msa(uint8_t *src,
         PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
         PCKEV_B2_UB(dst5, dst4, dst7, dst6, out2, out3);
 
-        ST6x4_UB(out0, out1, dst, dst_stride);
+        ST_W2(out0, 0, 2, dst, dst_stride);
+        ST_H2(out0, 2, 6, dst + 4, dst_stride);
+        ST_W2(out1, 0, 2, dst + 2 * dst_stride, dst_stride);
+        ST_H2(out1, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
         dst += (4 * dst_stride);
-        ST6x4_UB(out2, out3, dst, dst_stride);
+        ST_W2(out2, 0, 2, dst, dst_stride);
+        ST_H2(out2, 2, 6, dst + 4, dst_stride);
+        ST_W2(out3, 0, 2, dst + 2 * dst_stride, dst_stride);
+        ST_H2(out3, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -207,7 +213,7 @@  static void hevc_uniwgt_copy_8w_msa(uint8_t *src,
         HEVC_UNIW_RND_CLIP2_MAX_SATU_H(dst0, dst1, weight_vec, offset_vec,
                                        rnd_vec, dst0, dst1);
         out0 = (v16u8) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-        ST8x2_UB(out0, dst, dst_stride);
+        ST_D2(out0, 0, 1, dst, dst_stride);
     } else if (4 == height) {
         LD4(src, src_stride, tp0, tp1, tp2, tp3);
         INSERT_D2_SB(tp0, tp1, src0);
@@ -219,7 +225,7 @@  static void hevc_uniwgt_copy_8w_msa(uint8_t *src,
                                        offset_vec, rnd_vec, dst0, dst1, dst2,
                                        dst3);
         PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
     } else if (6 == height) {
         LD4(src, src_stride, tp0, tp1, tp2, tp3);
         src += 4 * src_stride;
@@ -238,9 +244,8 @@  static void hevc_uniwgt_copy_8w_msa(uint8_t *src,
         HEVC_UNIW_RND_CLIP2_MAX_SATU_H(dst4, dst5, weight_vec, offset_vec,
                                        rnd_vec, dst4, dst5);
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
-        ST8x4_UB(out0, out1, dst, dst_stride);
-        dst += (4 * dst_stride);
-        ST8x2_UB(out2, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+        ST_D2(out2, 0, 1, dst + 4 * dst_stride, dst_stride);
     } else if (0 == height % 8) {
         for (loop_cnt = (height >> 3); loop_cnt--;) {
             LD4(src, src_stride, tp0, tp1, tp2, tp3);
@@ -266,10 +271,9 @@  static void hevc_uniwgt_copy_8w_msa(uint8_t *src,
                                            dst6, dst7);
             PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
             PCKEV_B2_UB(dst5, dst4, dst7, dst6, out2, out3);
-            ST8x4_UB(out0, out1, dst, dst_stride);
-            dst += (4 * dst_stride);
-            ST8x4_UB(out2, out3, dst, dst_stride);
-            dst += (4 * dst_stride);
+            ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1,
+                  dst, dst_stride);
+            dst += (8 * dst_stride);
         }
     }
 }
@@ -313,7 +317,8 @@  static void hevc_uniwgt_copy_12w_msa(uint8_t *src,
                                        rnd_vec, dst4, dst5);
 
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
-        ST12x4_UB(out0, out1, out2, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(out2, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -409,7 +414,7 @@  static void hevc_uniwgt_copy_24w_msa(uint8_t *src,
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
         PCKEV_B3_UB(dst7, dst6, dst9, dst8, dst11, dst10, out3, out4, out5);
         ST_UB4(out0, out1, out3, out4, dst, dst_stride);
-        ST8x4_UB(out2, out5, dst + 16, dst_stride);
+        ST_D4(out2, out5, 0, 1, 0, 1, dst + 16, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -651,7 +656,7 @@  static void hevc_hz_uniwgt_8t_4w_msa(uint8_t *src,
                                        dst3);
 
         PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-        ST4x8_UB(out0, out1, dst, dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
     }
 }
@@ -729,7 +734,7 @@  static void hevc_hz_uniwgt_8t_8w_msa(uint8_t *src,
                                        dst3);
 
         PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -822,8 +827,8 @@  static void hevc_hz_uniwgt_8t_12w_msa(uint8_t *src,
                                        rnd_vec, dst4, dst5);
 
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
-        ST8x4_UB(out0, out1, dst, dst_stride);
-        ST4x4_UB(out2, out2, 0, 1, 2, 3, dst + 8, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(out2, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -994,7 +999,7 @@  static void hevc_hz_uniwgt_8t_24w_msa(uint8_t *src,
 
         PCKEV_B3_UB(dst1, dst0, dst4, dst3, dst5, dst2, out0, out1, out2);
         ST_UB2(out0, out1, dst, dst_stride);
-        ST8x2_UB(out2, dst + 16, dst_stride);
+        ST_D2(out2, 0, 1, dst + 16, dst_stride);
         dst += (2 * dst_stride);
     }
 }
@@ -1368,7 +1373,7 @@  static void hevc_vt_uniwgt_8t_4w_msa(uint8_t *src,
                                        dst3);
 
         PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-        ST4x8_UB(out0, out1, dst, dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
 
         src2110 = src10998;
@@ -1444,7 +1449,7 @@  static void hevc_vt_uniwgt_8t_8w_msa(uint8_t *src,
                                        dst3);
 
         PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         src10_r = src54_r;
@@ -1543,8 +1548,8 @@  static void hevc_vt_uniwgt_8t_12w_msa(uint8_t *src,
                                        rnd_vec, dst4, dst5);
 
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
-        ST8x4_UB(out0, out1, dst, dst_stride);
-        ST4x4_UB(out2, out2, 0, 1, 2, 3, dst + 8, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(out2, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
 
         src10_r = src54_r;
@@ -1861,7 +1866,7 @@  static void hevc_hv_uniwgt_8t_4w_msa(uint8_t *src,
         CLIP_SW4_0_255_MAX_SATU(dst0_r, dst1_r, dst2_r, dst3_r);
         PCKEV_H2_SW(dst1_r, dst0_r, dst3_r, dst2_r, dst0_r, dst1_r);
         out = (v16u8) __msa_pckev_b((v16i8) dst1_r, (v16i8) dst0_r);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         dst10_r = dst54_r;
@@ -2013,7 +2018,7 @@  static void hevc_hv_uniwgt_8t_8multx2mult_msa(uint8_t *src,
 
             PCKEV_H2_SW(dst0_l, dst0_r, dst1_l, dst1_r, dst0_r, dst1_r);
             dst0_r = (v4i32) __msa_pckev_b((v16i8) dst1_r, (v16i8) dst0_r);
-            ST8x2_UB(dst0_r, dst_tmp, dst_stride);
+            ST_D2(dst0_r, 0, 1, dst_tmp, dst_stride);
             dst_tmp += (2 * dst_stride);
 
             dst10_r = dst32_r;
@@ -2163,7 +2168,7 @@  static void hevc_hv_uniwgt_8t_12w_msa(uint8_t *src,
         CLIP_SW2_0_255_MAX_SATU(dst0_r, dst0_l);
         dst0_r = (v4i32) __msa_pckev_h((v8i16) dst0_l, (v8i16) dst0_r);
         out = (v16u8) __msa_pckev_b((v16i8) dst0_r, (v16i8) dst0_r);
-        ST8x1_UB(out, dst_tmp);
+        ST_D1(out, 0, dst_tmp);
         dst_tmp += dst_stride;
 
         dst0 = dst1;
@@ -2244,7 +2249,7 @@  static void hevc_hv_uniwgt_8t_12w_msa(uint8_t *src,
         CLIP_SW4_0_255_MAX_SATU(dst0_r, dst1_r, dst2_r, dst3_r);
         PCKEV_H2_SW(dst1_r, dst0_r, dst3_r, dst2_r, dst0_r, dst1_r);
         out = (v16u8) __msa_pckev_b((v16i8) dst1_r, (v16i8) dst0_r);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         dst10_r = dst54_r;
@@ -2391,7 +2396,7 @@  static void hevc_hz_uniwgt_4t_4x2_msa(uint8_t *src,
     dst0 = __msa_adds_s_h(dst0, offset_vec);
     dst0 = CLIP_SH_0_255_MAX_SATU(dst0);
     out = (v16u8) __msa_pckev_b((v16i8) dst0, (v16i8) dst0);
-    ST4x2_UB(out, dst, dst_stride);
+    ST_W2(out, 0, 1, dst, dst_stride);
     dst += (4 * dst_stride);
 }
 
@@ -2448,7 +2453,7 @@  static void hevc_hz_uniwgt_4t_4x4_msa(uint8_t *src,
                                    dst0, dst1);
 
     out = (v16u8) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
     dst += (4 * dst_stride);
 }
 
@@ -2515,7 +2520,7 @@  static void hevc_hz_uniwgt_4t_4x8multiple_msa(uint8_t *src,
                                        dst0, dst1, dst2, dst3);
 
         PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-        ST4x8_UB(out0, out1, dst, dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
     }
 }
@@ -2613,9 +2618,15 @@  static void hevc_hz_uniwgt_4t_6w_msa(uint8_t *src,
 
     PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
     PCKEV_B2_UB(dst5, dst4, dst7, dst6, out2, out3);
-    ST6x4_UB(out0, out1, dst, dst_stride);
+    ST_W2(out0, 0, 2, dst, dst_stride);
+    ST_H2(out0, 2, 6, dst + 4, dst_stride);
+    ST_W2(out1, 0, 2, dst + 2 * dst_stride, dst_stride);
+    ST_H2(out1, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
     dst += (4 * dst_stride);
-    ST6x4_UB(out2, out3, dst, dst_stride);
+    ST_W2(out2, 0, 2, dst, dst_stride);
+    ST_H2(out2, 2, 6, dst + 4, dst_stride);
+    ST_W2(out3, 0, 2, dst + 2 * dst_stride, dst_stride);
+    ST_H2(out3, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
 }
 
 static void hevc_hz_uniwgt_4t_8x2_msa(uint8_t *src,
@@ -2670,7 +2681,7 @@  static void hevc_hz_uniwgt_4t_8x2_msa(uint8_t *src,
                                    dst0, dst1);
 
     out = (v16u8) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-    ST8x2_UB(out, dst, dst_stride);
+    ST_D2(out, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hz_uniwgt_4t_8x4_msa(uint8_t *src,
@@ -2727,7 +2738,7 @@  static void hevc_hz_uniwgt_4t_8x4_msa(uint8_t *src,
                                    dst0, dst1, dst2, dst3);
 
     PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-    ST8x4_UB(out0, out1, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hz_uniwgt_4t_8x6_msa(uint8_t *src,
@@ -2796,9 +2807,8 @@  static void hevc_hz_uniwgt_4t_8x6_msa(uint8_t *src,
                                    dst4, dst5);
 
     PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
-    ST8x4_UB(out0, out1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST8x2_UB(out2, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+    ST_D2(out2, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void hevc_hz_uniwgt_4t_8x8multiple_msa(uint8_t *src,
@@ -2876,7 +2886,7 @@  static void hevc_hz_uniwgt_4t_8x8multiple_msa(uint8_t *src,
 
         PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
         PCKEV_B2_UB(dst5, dst4, dst7, dst6, out2, out3);
-        ST8x8_UB(out0, out1, out2, out3, dst, dst_stride);
+        ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride);
         dst += (8 * dst_stride);
     }
 }
@@ -2981,7 +2991,8 @@  static void hevc_hz_uniwgt_4t_12w_msa(uint8_t *src,
                                        rnd_vec, dst4, dst5);
 
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
-        ST12x4_UB(out0, out1, out2, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(out2, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -3142,7 +3153,7 @@  static void hevc_hz_uniwgt_4t_24w_msa(uint8_t *src,
 
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
         ST_UB2(out0, out1, dst, dst_stride);
-        ST8x2_UB(out2, dst + 16, dst_stride);
+        ST_D2(out2, 0, 1, dst + 16, dst_stride);
         dst += (2 * dst_stride);
     }
 }
@@ -3286,7 +3297,7 @@  static void hevc_vt_uniwgt_4t_4x2_msa(uint8_t *src,
     dst0 = __msa_adds_s_h(dst0, offset_vec);
     dst0 = CLIP_SH_0_255_MAX_SATU(dst0);
     out = (v16u8) __msa_pckev_b((v16i8) dst0, (v16i8) dst0);
-    ST4x2_UB(out, dst, dst_stride);
+    ST_W2(out, 0, 1, dst, dst_stride);
 }
 
 static void hevc_vt_uniwgt_4t_4x4_msa(uint8_t *src,
@@ -3340,7 +3351,7 @@  static void hevc_vt_uniwgt_4t_4x4_msa(uint8_t *src,
                                    dst0, dst1);
 
     out = (v16u8) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void hevc_vt_uniwgt_4t_4x8multiple_msa(uint8_t *src,
@@ -3411,7 +3422,7 @@  static void hevc_vt_uniwgt_4t_4x8multiple_msa(uint8_t *src,
                                        dst0, dst1, dst2, dst3);
 
         PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-        ST4x8_UB(out0, out1, dst, dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
 
         src2 = src10;
@@ -3509,9 +3520,15 @@  static void hevc_vt_uniwgt_4t_6w_msa(uint8_t *src,
 
     PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
     PCKEV_B2_UB(dst5, dst4, dst7, dst6, out2, out3);
-    ST6x4_UB(out0, out1, dst, dst_stride);
+    ST_W2(out0, 0, 2, dst, dst_stride);
+    ST_H2(out0, 2, 6, dst + 4, dst_stride);
+    ST_W2(out1, 0, 2, dst + 2 * dst_stride, dst_stride);
+    ST_H2(out1, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
     dst += (4 * dst_stride);
-    ST6x4_UB(out2, out3, dst, dst_stride);
+    ST_W2(out2, 0, 2, dst, dst_stride);
+    ST_H2(out2, 2, 6, dst + 4, dst_stride);
+    ST_W2(out3, 0, 2, dst + 2 * dst_stride, dst_stride);
+    ST_H2(out3, 2, 6, dst + 2 * dst_stride + 4, dst_stride);
 }
 
 static void hevc_vt_uniwgt_4t_8x2_msa(uint8_t *src,
@@ -3562,7 +3579,7 @@  static void hevc_vt_uniwgt_4t_8x2_msa(uint8_t *src,
                                    dst0, dst1);
 
     out = (v16u8) __msa_pckev_b((v16i8) dst1, (v16i8) dst0);
-    ST8x2_UB(out, dst, dst_stride);
+    ST_D2(out, 0, 1, dst, dst_stride);
 }
 
 static void hevc_vt_uniwgt_4t_8x4_msa(uint8_t *src,
@@ -3617,7 +3634,7 @@  static void hevc_vt_uniwgt_4t_8x4_msa(uint8_t *src,
                                    offset_vec, rnd_vec, dst0, dst1, dst2,
                                    dst3);
     PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
-    ST8x4_UB(out0, out1, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hevc_vt_uniwgt_4t_8x6_msa(uint8_t *src,
@@ -3679,9 +3696,8 @@  static void hevc_vt_uniwgt_4t_8x6_msa(uint8_t *src,
     HEVC_UNIW_RND_CLIP2_MAX_SATU_H(dst4, dst5, weight_vec, offset_vec, rnd_vec,
                                    dst4, dst5);
     PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
-    ST8x4_UB(out0, out1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST8x2_UB(out2, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+    ST_D2(out2, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void hevc_vt_uniwgt_4t_8x8mult_msa(uint8_t *src,
@@ -3754,7 +3770,7 @@  static void hevc_vt_uniwgt_4t_8x8mult_msa(uint8_t *src,
                                        dst7);
         PCKEV_B2_UB(dst1, dst0, dst3, dst2, out0, out1);
         PCKEV_B2_UB(dst5, dst4, dst7, dst6, out2, out3);
-        ST8x8_UB(out0, out1, out2, out3, dst, dst_stride);
+        ST_D8(out0, out1, out2, out3, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride);
         dst += (8 * dst_stride);
 
         src2 = src10;
@@ -3861,7 +3877,8 @@  static void hevc_vt_uniwgt_4t_12w_msa(uint8_t *src,
         HEVC_UNIW_RND_CLIP2_MAX_SATU_H(dst4, dst5, weight_vec, offset_vec,
                                        rnd_vec, dst4, dst5);
         PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2);
-        ST12x4_UB(out0, out1, out2, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(out2, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
 
         ILVRL_B2_SB(src7, src6, src76_r, src76_l);
@@ -3882,7 +3899,8 @@  static void hevc_vt_uniwgt_4t_12w_msa(uint8_t *src,
         HEVC_UNIW_RND_CLIP2_MAX_SATU_H(dst10, dst11, weight_vec, offset_vec,
                                        rnd_vec, dst10, dst11);
         PCKEV_B3_UB(dst7, dst6, dst9, dst8, dst11, dst10, out3, out4, out5);
-        ST12x4_UB(out3, out4, out5, dst, dst_stride);
+        ST_D4(out3, out4, 0, 1, 0, 1, dst, dst_stride);
+        ST_W4(out5, 0, 1, 2, 3, dst + 8, dst_stride);
         dst += (4 * dst_stride);
 
         src2 = src10;
@@ -4062,7 +4080,7 @@  static void hevc_vt_uniwgt_4t_24w_msa(uint8_t *src,
                     out2, out3);
         PCKEV_B2_UB(dst9, dst8, dst11, dst10, out4, out5);
         ST_UB4(out0, out1, out2, out3, dst, dst_stride);
-        ST8x4_UB(out4, out5, dst + 16, dst_stride);
+        ST_D4(out4, out5, 0, 1, 0, 1, dst + 16, dst_stride);
         dst += (4 * dst_stride);
 
         src2 = src6;
@@ -4231,7 +4249,7 @@  static void hevc_hv_uniwgt_4t_4x2_msa(uint8_t *src,
     tmp += offset_vec;
     tmp = CLIP_SH_0_255_MAX_SATU(tmp);
     out = (v16u8) __msa_pckev_b((v16i8) tmp, (v16i8) tmp);
-    ST4x2_UB(out, dst, dst_stride);
+    ST_W2(out, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hv_uniwgt_4t_4x4_msa(uint8_t *src,
@@ -4300,7 +4318,7 @@  static void hevc_hv_uniwgt_4t_4x4_msa(uint8_t *src,
     ADD2(tmp0, offset_vec, tmp1, offset_vec, tmp0, tmp1);
     CLIP_SH2_0_255_MAX_SATU(tmp0, tmp1);
     out = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void hevc_hv_uniwgt_4t_4multx8mult_msa(uint8_t *src,
@@ -4401,7 +4419,7 @@  static void hevc_hv_uniwgt_4t_4multx8mult_msa(uint8_t *src,
         ADD2(tmp2, offset_vec, tmp3, offset_vec, tmp2, tmp3);
         CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
         PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST4x8_UB(out0, out1, dst, dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
 
         dst10_r = dst98_r;
@@ -4559,10 +4577,8 @@  static void hevc_hv_uniwgt_4t_6w_msa(uint8_t *src,
     CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
     CLIP_SH2_0_255_MAX_SATU(tmp4, tmp5);
     PCKEV_B3_UB(tmp1, tmp0, tmp3, tmp2, tmp5, tmp4, out0, out1, out2);
-    ST4x8_UB(out0, out1, dst, dst_stride);
-    ST2x4_UB(out2, 0, dst + 4, dst_stride);
-    dst += 4 * dst_stride;
-    ST2x4_UB(out2, 4, dst + 4, dst_stride);
+    ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
+    ST_H8(out2, 0, 1, 2, 3, 4, 5, 6, 7, dst + 4, dst_stride);
 }
 
 static void hevc_hv_uniwgt_4t_8x2_msa(uint8_t *src,
@@ -4638,7 +4654,7 @@  static void hevc_hv_uniwgt_4t_8x2_msa(uint8_t *src,
     ADD2(tmp0, offset_vec, tmp1, offset_vec, tmp0, tmp1);
     CLIP_SH2_0_255_MAX_SATU(tmp0, tmp1);
     out = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-    ST8x2_UB(out, dst, dst_stride);
+    ST_D2(out, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hv_uniwgt_4t_8multx4_msa(uint8_t *src,
@@ -4731,7 +4747,7 @@  static void hevc_hv_uniwgt_4t_8multx4_msa(uint8_t *src,
         ADD2(tmp2, offset_vec, tmp3, offset_vec, tmp2, tmp3);
         CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
         PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
         dst += 8;
     }
 }
@@ -4848,9 +4864,8 @@  static void hevc_hv_uniwgt_4t_8x6_msa(uint8_t *src,
     CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
     CLIP_SH2_0_255_MAX_SATU(tmp4, tmp5);
     PCKEV_B3_UB(tmp1, tmp0, tmp3, tmp2, tmp5, tmp4, out0, out1, out2);
-    ST8x4_UB(out0, out1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST8x2_UB(out2, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
+    ST_D2(out2, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void hevc_hv_uniwgt_4t_8multx4mult_msa(uint8_t *src,
@@ -4960,7 +4975,7 @@  static void hevc_hv_uniwgt_4t_8multx4mult_msa(uint8_t *src,
             ADD2(tmp2, offset_vec, tmp3, offset_vec, tmp2, tmp3);
             CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
             PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-            ST8x4_UB(out0, out1, dst_tmp, dst_stride);
+            ST_D4(out0, out1, 0, 1, 0, 1, dst_tmp, dst_stride);
             dst_tmp += (4 * dst_stride);
 
             dst10_r = dst54_r;
@@ -5107,7 +5122,7 @@  static void hevc_hv_uniwgt_4t_12w_msa(uint8_t *src,
         ADD2(tmp2, offset_vec, tmp3, offset_vec, tmp2, tmp3);
         CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
         PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST8x4_UB(out0, out1, dst_tmp, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst_tmp, dst_stride);
         dst_tmp += (4 * dst_stride);
 
         dst10_r = dst54_r;
@@ -5174,7 +5189,7 @@  static void hevc_hv_uniwgt_4t_12w_msa(uint8_t *src,
         ADD2(tmp2, offset_vec, tmp3, offset_vec, tmp2, tmp3);
         CLIP_SH4_0_255_MAX_SATU(tmp0, tmp1, tmp2, tmp3);
         PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST4x8_UB(out0, out1, dst, dst_stride);
+        ST_W8(out0, out1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
         dst += (8 * dst_stride);
 
         dst10_r = dst98_r;
diff --git a/libavcodec/mips/hevcdsp_msa.c b/libavcodec/mips/hevcdsp_msa.c
index 81db62b..2c57ec8 100644
--- a/libavcodec/mips/hevcdsp_msa.c
+++ b/libavcodec/mips/hevcdsp_msa.c
@@ -44,7 +44,7 @@  static void hevc_copy_4w_msa(uint8_t *src, int32_t src_stride,
         src0 = (v16i8) __msa_ilvr_w((v4i32) src1, (v4i32) src0);
         in0 = (v8i16) __msa_ilvr_b(zero, src0);
         in0 <<= 6;
-        ST8x2_UB(in0, dst, 2 * dst_stride);
+        ST_D2(in0, 0, 1, dst, dst_stride);
     } else if (4 == height) {
         v16i8 src0, src1, src2, src3;
         v8i16 in0, in1;
@@ -55,7 +55,7 @@  static void hevc_copy_4w_msa(uint8_t *src, int32_t src_stride,
         ILVR_B2_SH(zero, src0, zero, src1, in0, in1);
         in0 <<= 6;
         in1 <<= 6;
-        ST8x4_UB(in0, in1, dst, 2 * dst_stride);
+        ST_D4(in0, in1, 0, 1, 0, 1, dst, dst_stride);
     } else if (0 == height % 8) {
         v16i8 src0, src1, src2, src3, src4, src5, src6, src7;
         v8i16 in0, in1, in2, in3;
@@ -71,7 +71,7 @@  static void hevc_copy_4w_msa(uint8_t *src, int32_t src_stride,
             ILVR_B4_SH(zero, src0, zero, src1, zero, src2, zero, src3,
                        in0, in1, in2, in3);
             SLLI_4V(in0, in1, in2, in3, 6);
-            ST8x8_UB(in0, in1, in2, in3, dst, 2 * dst_stride);
+            ST_D8(in0, in1, in2, in3, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride);
             dst += (8 * dst_stride);
         }
     }
@@ -183,7 +183,7 @@  static void hevc_copy_12w_msa(uint8_t *src, int32_t src_stride,
         in0 <<= 6;
         in1 <<= 6;
         ST_SH4(in0_r, in1_r, in2_r, in3_r, dst, dst_stride);
-        ST8x4_UB(in0, in1, dst + 8, 2 * dst_stride);
+        ST_D4(in0, in1, 0, 1, 0, 1, dst + 8, dst_stride);
         dst += (4 * dst_stride);
 
         ILVR_B4_SH(zero, src4, zero, src5, zero, src6, zero, src7,
@@ -194,7 +194,7 @@  static void hevc_copy_12w_msa(uint8_t *src, int32_t src_stride,
         in0 <<= 6;
         in1 <<= 6;
         ST_SH4(in0_r, in1_r, in2_r, in3_r, dst, dst_stride);
-        ST8x4_UB(in0, in1, dst + 8, 2 * dst_stride);
+        ST_D4(in0, in1, 0, 1, 0, 1, dst + 8, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -495,7 +495,7 @@  static void hevc_hz_8t_4w_msa(uint8_t *src, int32_t src_stride,
         DPADD_SB4_SH(vec0, vec1, vec2, vec3, filt0, filt1, filt2, filt3,
                      dst3, dst3, dst3, dst3);
 
-        ST8x8_UB(dst0, dst1, dst2, dst3, dst, 2 * dst_stride);
+        ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride);
         dst += (8 * dst_stride);
     }
 }
@@ -1047,7 +1047,7 @@  static void hevc_vt_8t_4w_msa(uint8_t *src, int32_t src_stride,
         DPADD_SB4_SH(src8776, src10998, src12111110, src14131312,
                      filt0, filt1, filt2, filt3, dst76, dst76, dst76, dst76);
 
-        ST8x8_UB(dst10, dst32, dst54, dst76, dst, 2 * dst_stride);
+        ST_D8(dst10, dst32, dst54, dst76, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride);
         dst += (8 * dst_stride);
 
         src2110 = src10998;
@@ -1191,7 +1191,7 @@  static void hevc_vt_8t_12w_msa(uint8_t *src, int32_t src_stride,
                      dst1_l, dst1_l, dst1_l, dst1_l);
 
         ST_SH4(dst0_r, dst1_r, dst2_r, dst3_r, dst, dst_stride);
-        ST8x4_UB(dst0_l, dst1_l, dst + 8, 2 * dst_stride);
+        ST_D4(dst0_l, dst1_l, 0, 1, 0, 1, dst + 8, dst_stride);
         dst += (4 * dst_stride);
 
         src10_r = src54_r;
@@ -1363,7 +1363,6 @@  static void hevc_hv_8t_4w_msa(uint8_t *src, int32_t src_stride,
                               int32_t height)
 {
     uint32_t loop_cnt;
-    int32_t dst_stride_in_bytes = 2 * dst_stride;
     v16i8 src0, src1, src2, src3, src4, src5, src6, src7, src8, src9, src10;
     v8i16 filt0, filt1, filt2, filt3;
     v8i16 filt_h0, filt_h1, filt_h2, filt_h3;
@@ -1452,7 +1451,7 @@  static void hevc_hv_8t_4w_msa(uint8_t *src, int32_t src_stride,
                                 filt_h0, filt_h1, filt_h2, filt_h3);
         SRA_4V(dst0_r, dst1_r, dst2_r, dst3_r, 6);
         PCKEV_H2_SW(dst1_r, dst0_r, dst3_r, dst2_r, dst0_r, dst2_r);
-        ST8x4_UB(dst0_r, dst2_r, dst, dst_stride_in_bytes);
+        ST_D4(dst0_r, dst2_r, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         dst10_r = dst54_r;
@@ -1607,7 +1606,6 @@  static void hevc_hv_8t_12w_msa(uint8_t *src, int32_t src_stride,
                                int32_t height)
 {
     uint32_t loop_cnt;
-    int32_t dst_stride_in_bytes = 2 * dst_stride;
     uint8_t *src_tmp;
     int16_t *dst_tmp;
     v16i8 src0, src1, src2, src3, src4, src5, src6, src7, src8, src9, src10;
@@ -1784,7 +1782,7 @@  static void hevc_hv_8t_12w_msa(uint8_t *src, int32_t src_stride,
                                 filt_h1, filt_h2, filt_h3);
         SRA_4V(dst0_r, dst1_r, dst2_r, dst3_r, 6);
         PCKEV_H2_SW(dst1_r, dst0_r, dst3_r, dst2_r, dst0_r, dst2_r);
-        ST8x4_UB(dst0_r, dst2_r, dst, dst_stride_in_bytes);
+        ST_D4(dst0_r, dst2_r, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         dst10_r = dst54_r;
@@ -1872,7 +1870,7 @@  static void hevc_hz_4t_4x2_msa(uint8_t *src,
     dst0 = const_vec;
     DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst0, dst0);
 
-    ST8x2_UB(dst0, dst, 2 * dst_stride);
+    ST_D2(dst0, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hz_4t_4x4_msa(uint8_t *src,
@@ -1909,7 +1907,7 @@  static void hevc_hz_4t_4x4_msa(uint8_t *src,
     dst1 = const_vec;
     DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst1, dst1);
 
-    ST8x4_UB(dst0, dst1, dst, 2 * dst_stride);
+    ST_D4(dst0, dst1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hz_4t_4x8multiple_msa(uint8_t *src,
@@ -1956,7 +1954,7 @@  static void hevc_hz_4t_4x8multiple_msa(uint8_t *src,
         dst3 = const_vec;
         DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst3, dst3);
 
-        ST8x8_UB(dst0, dst1, dst2, dst3, dst, 2 * dst_stride);
+        ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride);
         dst += (8 * dst_stride);
     }
 }
@@ -2218,7 +2216,7 @@  static void hevc_hz_4t_12w_msa(uint8_t *src,
         DPADD_SB2_SH(vec0, vec1, filt0, filt1, dst5, dst5);
 
         ST_SH4(dst0, dst1, dst2, dst3, dst, dst_stride);
-        ST8x4_UB(dst4, dst5, dst + 8, 2 * dst_stride);
+        ST_D4(dst4, dst5, 0, 1, 0, 1, dst + 8, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -2478,7 +2476,7 @@  static void hevc_vt_4t_4x2_msa(uint8_t *src,
     dst10 = const_vec;
     DPADD_SB2_SH(src2110, src4332, filt0, filt1, dst10, dst10);
 
-    ST8x2_UB(dst10, dst, 2 * dst_stride);
+    ST_D2(dst10, 0, 1, dst, dst_stride);
 }
 
 static void hevc_vt_4t_4x4_msa(uint8_t *src,
@@ -2515,7 +2513,7 @@  static void hevc_vt_4t_4x4_msa(uint8_t *src,
     dst32 = const_vec;
     DPADD_SB2_SH(src4332, src6554, filt0, filt1, dst32, dst32);
 
-    ST8x4_UB(dst10, dst32, dst, 2 * dst_stride);
+    ST_D4(dst10, dst32, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hevc_vt_4t_4x8_msa(uint8_t *src,
@@ -2564,8 +2562,7 @@  static void hevc_vt_4t_4x8_msa(uint8_t *src,
     DPADD_SB2_SH(src4332, src6554, filt0, filt1, dst32, dst32);
     DPADD_SB2_SH(src6554, src8776, filt0, filt1, dst54, dst54);
     DPADD_SB2_SH(src8776, src10998, filt0, filt1, dst76, dst76);
-    ST8x8_UB(dst10, dst32, dst54, dst76, dst, 2 * dst_stride);
-    dst += (8 * dst_stride);
+    ST_D8(dst10, dst32, dst54, dst76, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hevc_vt_4t_4x16_msa(uint8_t *src, int32_t src_stride,
@@ -2610,7 +2607,7 @@  static void hevc_vt_4t_4x16_msa(uint8_t *src, int32_t src_stride,
     DPADD_SB2_SH(src4332, src6554, filt0, filt1, dst32, dst32);
     DPADD_SB2_SH(src6554, src8776, filt0, filt1, dst54, dst54);
     DPADD_SB2_SH(src8776, src10998, filt0, filt1, dst76, dst76);
-    ST8x8_UB(dst10, dst32, dst54, dst76, dst, 2 * dst_stride);
+    ST_D8(dst10, dst32, dst54, dst76, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride);
     dst += (8 * dst_stride);
 
     src2 = src10;
@@ -2635,8 +2632,7 @@  static void hevc_vt_4t_4x16_msa(uint8_t *src, int32_t src_stride,
     DPADD_SB2_SH(src4332, src6554, filt0, filt1, dst32, dst32);
     DPADD_SB2_SH(src6554, src8776, filt0, filt1, dst54, dst54);
     DPADD_SB2_SH(src8776, src10998, filt0, filt1, dst76, dst76);
-    ST8x8_UB(dst10, dst32, dst54, dst76, dst, 2 * dst_stride);
-    dst += (8 * dst_stride);
+    ST_D8(dst10, dst32, dst54, dst76, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hevc_vt_4t_4w_msa(uint8_t *src,
@@ -2955,7 +2951,7 @@  static void hevc_vt_4t_12w_msa(uint8_t *src,
         DPADD_SB2_SH(src4332, src6554, filt0, filt1, dst1_l, dst1_l);
 
         ST_SH4(dst0_r, dst1_r, dst2_r, dst3_r, dst, dst_stride);
-        ST8x4_UB(dst0_l, dst1_l, dst + 8, (2 * dst_stride));
+        ST_D4(dst0_l, dst1_l, 0, 1, 0, 1, dst + 8, dst_stride);
         dst += (4 * dst_stride);
 
         src2 = src6;
@@ -3243,7 +3239,6 @@  static void hevc_hv_4t_4x2_msa(uint8_t *src,
                                const int8_t *filter_x,
                                const int8_t *filter_y)
 {
-    int32_t dst_stride_in_bytes = 2 * dst_stride;
     v16i8 src0, src1, src2, src3, src4;
     v8i16 filt0, filt1;
     v8i16 filt_h0, filt_h1;
@@ -3288,7 +3283,7 @@  static void hevc_hv_4t_4x2_msa(uint8_t *src,
     dst0 >>= 6;
     dst1 >>= 6;
     dst0 = (v4i32) __msa_pckev_h((v8i16) dst1, (v8i16) dst0);
-    ST8x2_UB(dst0, dst, dst_stride_in_bytes);
+    ST_D2(dst0, 0, 1, dst, dst_stride);
 }
 
 static void hevc_hv_4t_4x4_msa(uint8_t *src,
@@ -3298,7 +3293,6 @@  static void hevc_hv_4t_4x4_msa(uint8_t *src,
                                const int8_t *filter_x,
                                const int8_t *filter_y)
 {
-    int32_t dst_stride_in_bytes = 2 * dst_stride;
     v16i8 src0, src1, src2, src3, src4, src5, src6;
     v8i16 filt0, filt1;
     v8i16 filt_h0, filt_h1;
@@ -3351,7 +3345,7 @@  static void hevc_hv_4t_4x4_msa(uint8_t *src,
     dst3 = HEVC_FILT_4TAP(dst43, dst65, filt_h0, filt_h1);
     SRA_4V(dst0, dst1, dst2, dst3, 6);
     PCKEV_H2_SW(dst1, dst0, dst3, dst2, dst0, dst2);
-    ST8x4_UB(dst0, dst2, dst, dst_stride_in_bytes);
+    ST_D4(dst0, dst2, 0, 1, 0, 1, dst, dst_stride);
 }
 
 
@@ -3442,7 +3436,7 @@  static void hevc_hv_4t_4multx8mult_msa(uint8_t *src,
         SRA_4V(dst4, dst5, dst6, dst7, 6);
         PCKEV_H4_SW(dst1, dst0, dst3, dst2, dst5, dst4, dst7, dst6,
                     dst0, dst1, dst2, dst3);
-        ST8x8_UB(dst0, dst1, dst2, dst3, dst, 2 * dst_stride);
+        ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride);
         dst += (8 * dst_stride);
 
         dst10_r = dst98_r;
@@ -3479,7 +3473,6 @@  static void hevc_hv_4t_6w_msa(uint8_t *src,
                               const int8_t *filter_y,
                               int32_t height)
 {
-    int32_t dst_stride_in_bytes = 2 * dst_stride;
     v16i8 src0, src1, src2, src3, src4, src5, src6, src7, src8, src9, src10;
     v8i16 filt0, filt1;
     v8i16 filt_h0, filt_h1;
@@ -3590,11 +3583,11 @@  static void hevc_hv_4t_6w_msa(uint8_t *src,
     PCKEV_H2_SH(dst1_r, dst0_r, dst3_r, dst2_r, tmp0, tmp1);
     PCKEV_H2_SH(dst5_r, dst4_r, dst7_r, dst6_r, tmp2, tmp3);
     PCKEV_H2_SH(dst1_l, dst0_l, dst3_l, dst2_l, tmp4, tmp5);
-    ST8x4_UB(tmp0, tmp1, dst, dst_stride_in_bytes);
-    ST4x4_UB(tmp4, tmp4, 0, 1, 2, 3, dst + 4, dst_stride_in_bytes);
+    ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
+    ST_W4(tmp4, 0, 1, 2, 3, dst + 4, dst_stride);
     dst += 4 * dst_stride;
-    ST8x4_UB(tmp2, tmp3, dst, dst_stride_in_bytes);
-    ST4x4_UB(tmp5, tmp5, 0, 1, 2, 3, dst + 4, dst_stride_in_bytes);
+    ST_D4(tmp2, tmp3, 0, 1, 0, 1, dst, dst_stride);
+    ST_W4(tmp5, 0, 1, 2, 3, dst + 4, dst_stride);
 }
 
 static void hevc_hv_4t_8x2_msa(uint8_t *src,
@@ -4164,7 +4157,7 @@  static void hevc_hv_4t_12w_msa(uint8_t *src,
         SRA_4V(tmp4, tmp5, tmp6, tmp7, 6);
         PCKEV_H4_SW(tmp1, tmp0, tmp3, tmp2, tmp5, tmp4, tmp7, tmp6, tmp0, tmp1,
                     tmp2, tmp3);
-        ST8x8_UB(tmp0, tmp1, tmp2, tmp3, dst, 2 * dst_stride);
+        ST_D8(tmp0, tmp1, tmp2, tmp3, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride);
         dst += (8 * dst_stride);
 
         dst10_r = dst98_r;
diff --git a/libavcodec/mips/hevcpred_msa.c b/libavcodec/mips/hevcpred_msa.c
index 963c64c..b8df089 100644
--- a/libavcodec/mips/hevcpred_msa.c
+++ b/libavcodec/mips/hevcpred_msa.c
@@ -589,7 +589,7 @@  static void hevc_intra_pred_plane_4x4_msa(const uint8_t *src_top,
     PCKEV_D2_SH(res1, res0, res3, res2, res0, res1);
     SRARI_H2_SH(res0, res1, 3);
     src_vec0 = __msa_pckev_b((v16i8) res1, (v16i8) res0);
-    ST4x4_UB(src_vec0, src_vec0, 0, 1, 2, 3, dst, stride);
+    ST_W4(src_vec0, 0, 1, 2, 3, dst, stride);
 }
 
 static void hevc_intra_pred_plane_8x8_msa(const uint8_t *src_top,
@@ -656,7 +656,8 @@  static void hevc_intra_pred_plane_8x8_msa(const uint8_t *src_top,
     PCKEV_B4_SB(res1, res0, res3, res2, res5, res4, res7, res6,
                 src_vec0, src_vec1, src_vec2, src_vec3);
 
-    ST8x8_UB(src_vec0, src_vec1, src_vec2, src_vec3, dst, stride);
+    ST_D8(src_vec0, src_vec1, src_vec2, src_vec3, 0, 1, 0, 1,
+          0, 1, 0, 1, dst, stride);
 }
 
 static void hevc_intra_pred_plane_16x16_msa(const uint8_t *src_top,
@@ -1007,7 +1008,7 @@  static void hevc_intra_pred_angular_upper_4width_msa(const uint8_t *src_top,
 
     SRARI_H2_SH(diff1, diff3, 5);
     dst_val0 = __msa_pckev_b((v16i8) diff3, (v16i8) diff1);
-    ST4x4_UB(dst_val0, dst_val0, 0, 1, 2, 3, dst, stride);
+    ST_W4(dst_val0, 0, 1, 2, 3, dst, stride);
 }
 
 static void hevc_intra_pred_angular_upper_8width_msa(const uint8_t *src_top,
@@ -1104,7 +1105,7 @@  static void hevc_intra_pred_angular_upper_8width_msa(const uint8_t *src_top,
 
         SRARI_H4_SH(diff1, diff3, diff5, diff7, 5);
         PCKEV_B2_UB(diff3, diff1, diff7, diff5, dst_val0, dst_val1);
-        ST8x4_UB(dst_val0, dst_val1, dst, stride);
+        ST_D4(dst_val0, dst_val1, 0, 1, 0, 1, dst, stride);
         dst += (4 * stride);
     }
 }
@@ -1425,9 +1426,8 @@  static void hevc_intra_pred_angular_lower_4width_msa(const uint8_t *src_top,
     dst_val0 = __msa_pckev_b((v16i8) diff2, (v16i8) diff2);
     dst_val1 = __msa_pckod_b((v16i8) diff2, (v16i8) diff2);
 
-    ST4x2_UB(dst_val0, dst, stride);
-    dst += (2 * stride);
-    ST4x2_UB(dst_val1, dst, stride);
+    ST_W2(dst_val0, 0, 1, dst, stride);
+    ST_W2(dst_val1, 0, 1, dst + 2 * stride, stride);
 }
 
 static void hevc_intra_pred_angular_lower_8width_msa(const uint8_t *src_top,
@@ -1526,7 +1526,7 @@  static void hevc_intra_pred_angular_lower_8width_msa(const uint8_t *src_top,
                     dst_val0, dst_val1, dst_val2, dst_val3);
         ILVR_B2_SH(dst_val1, dst_val0, dst_val3, dst_val2, diff0, diff1);
         ILVRL_H2_SH(diff1, diff0, diff3, diff4);
-        ST4x8_UB(diff3, diff4, dst_org, stride);
+        ST_W8(diff3, diff4, 0, 1, 2, 3, 0, 1, 2, 3, dst_org, stride);
         dst += 4;
     }
 }
@@ -1640,9 +1640,9 @@  static void hevc_intra_pred_angular_lower_16width_msa(const uint8_t *src_top,
         ILVL_B2_SH(dst_val1, dst_val0, dst_val3, dst_val2, diff2, diff3);
         ILVRL_H2_SH(diff1, diff0, diff4, diff5);
         ILVRL_H2_SH(diff3, diff2, diff6, diff7);
-        ST4x8_UB(diff4, diff5, dst_org, stride);
+        ST_W8(diff4, diff5, 0, 1, 2, 3, 0, 1, 2, 3, dst_org, stride);
         dst_org += (8 * stride);
-        ST4x8_UB(diff6, diff7, dst_org, stride);
+        ST_W8(diff6, diff7, 0, 1, 2, 3, 0, 1, 2, 3, dst_org, stride);
         dst += 4;
     }
 }
@@ -1746,23 +1746,14 @@  static void hevc_intra_pred_angular_lower_32width_msa(const uint8_t *src_top,
         ILVRL_B2_SH(dst_val2, dst_val0, diff0, diff1);
         ILVRL_B2_SH(dst_val3, dst_val1, diff2, diff3);
 
-        ST2x4_UB(diff0, 0, dst_org, stride);
-        dst_org += (4 * stride);
-        ST2x4_UB(diff0, 4, dst_org, stride);
-        dst_org += (4 * stride);
-        ST2x4_UB(diff1, 0, dst_org, stride);
-        dst_org += (4 * stride);
-        ST2x4_UB(diff1, 4, dst_org, stride);
-        dst_org += (4 * stride);
-
-        ST2x4_UB(diff2, 0, dst_org, stride);
-        dst_org += (4 * stride);
-        ST2x4_UB(diff2, 4, dst_org, stride);
-        dst_org += (4 * stride);
-        ST2x4_UB(diff3, 0, dst_org, stride);
-        dst_org += (4 * stride);
-        ST2x4_UB(diff3, 4, dst_org, stride);
-        dst_org += (4 * stride);
+        ST_H8(diff0, 0, 1, 2, 3, 4, 5, 6, 7, dst_org, stride)
+        dst_org += (8 * stride);
+        ST_H8(diff1, 0, 1, 2, 3, 4, 5, 6, 7, dst_org, stride)
+        dst_org += (8 * stride);
+        ST_H8(diff2, 0, 1, 2, 3, 4, 5, 6, 7, dst_org, stride)
+        dst_org += (8 * stride);
+        ST_H8(diff3, 0, 1, 2, 3, 4, 5, 6, 7, dst_org, stride)
+        dst_org += (8 * stride);
 
         dst += 2;
     }
diff --git a/libavcodec/mips/hpeldsp_msa.c b/libavcodec/mips/hpeldsp_msa.c
index 40a0dca..ad92f8f 100644
--- a/libavcodec/mips/hpeldsp_msa.c
+++ b/libavcodec/mips/hpeldsp_msa.c
@@ -49,7 +49,7 @@ 
     PCKEV_B2_UB(in2, in1, in4, in3, tmp0_m, tmp1_m);                    \
     PCKEV_D2_UB(dst1, dst0, dst3, dst2, tmp2_m, tmp3_m);                \
     AVER_UB2_UB(tmp0_m, tmp2_m, tmp1_m, tmp3_m, tmp0_m, tmp1_m);        \
-    ST8x4_UB(tmp0_m, tmp1_m, pdst_m, stride);                           \
+    ST_D4(tmp0_m, tmp1_m, 0, 1, 0, 1, pdst_m, stride);                  \
 }
 
 static void common_hz_bil_4w_msa(const uint8_t *src, int32_t src_stride,
@@ -584,7 +584,7 @@  static void common_hv_bil_8w_msa(const uint8_t *src, int32_t src_stride,
              sum0, sum1, sum2, sum3);
         SRARI_H4_UH(sum0, sum1, sum2, sum3, 2);
         PCKEV_B2_SB(sum1, sum0, sum3, sum2, src0, src1);
-        ST8x4_UB(src0, src1, dst, dst_stride);
+        ST_D4(src0, src1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
         src0 = src4;
     }
@@ -689,9 +689,9 @@  static void common_hv_bil_no_rnd_8x8_msa(const uint8_t *src, int32_t src_stride,
     SRA_4V(sum0, sum1, sum2, sum3, 2);
     SRA_4V(sum4, sum5, sum6, sum7, 2);
     PCKEV_B2_SB(sum1, sum0, sum3, sum2, out0, out1);
-    ST8x4_UB(out0, out1, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
     PCKEV_B2_SB(sum5, sum4, sum7, sum6, out0, out1);
-    ST8x4_UB(out0, out1, dst + 4 * dst_stride, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void common_hv_bil_no_rnd_4x8_msa(const uint8_t *src, int32_t src_stride,
@@ -723,7 +723,7 @@  static void common_hv_bil_no_rnd_4x8_msa(const uint8_t *src, int32_t src_stride,
 
     SRA_4V(sum0, sum1, sum2, sum3, 2);
     PCKEV_B2_SB(sum1, sum0, sum3, sum2, out0, out1);
-    ST8x4_UB(out0, out1, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void common_hv_bil_no_rnd_16x16_msa(const uint8_t *src,
diff --git a/libavcodec/mips/qpeldsp_msa.c b/libavcodec/mips/qpeldsp_msa.c
index 4710b3f..fba42b30 100644
--- a/libavcodec/mips/qpeldsp_msa.c
+++ b/libavcodec/mips/qpeldsp_msa.c
@@ -334,7 +334,7 @@  static void horiz_mc_qpel_aver_src0_8width_msa(const uint8_t *src,
         inp0 = (v16u8) __msa_insve_d((v2i64) inp0, 1, (v2i64) inp1);
         inp2 = (v16u8) __msa_insve_d((v2i64) inp2, 1, (v2i64) inp3);
         AVER_UB2_UB(inp0, res0, inp2, res1, res0, res1);
-        ST8x4_UB(res0, res1, dst, dst_stride);
+        ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -409,7 +409,7 @@  static void horiz_mc_qpel_8width_msa(const uint8_t *src,
         res1 = APPLY_HORIZ_QPEL_FILTER_8BYTE(inp2, inp3,
                                              mask0, mask1, mask2, mask3,
                                              const20, const6, const3);
-        ST8x4_UB(res0, res1, dst, dst_stride);
+        ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -485,7 +485,7 @@  static void horiz_mc_qpel_aver_src1_8width_msa(const uint8_t *src,
         inp0 = (v16u8) __msa_insve_d((v2i64) inp0, 1, (v2i64) inp1);
         inp2 = (v16u8) __msa_insve_d((v2i64) inp2, 1, (v2i64) inp3);
         AVER_UB2_UB(inp0, res0, inp2, res1, res0, res1);
-        ST8x4_UB(res0, res1, dst, dst_stride);
+        ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -564,7 +564,7 @@  static void horiz_mc_qpel_no_rnd_aver_src0_8width_msa(const uint8_t *src,
         inp2 = (v16u8) __msa_insve_d((v2i64) inp2, 1, (v2i64) inp3);
         res0 = __msa_ave_u_b(inp0, res0);
         res1 = __msa_ave_u_b(inp2, res1);
-        ST8x4_UB(res0, res1, dst, dst_stride);
+        ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -639,7 +639,7 @@  static void horiz_mc_qpel_no_rnd_8width_msa(const uint8_t *src,
         res1 = APPLY_HORIZ_QPEL_NO_ROUND_FILTER_8BYTE(inp2, inp3, mask0, mask1,
                                                       mask2, mask3, const20,
                                                       const6, const3);
-        ST8x4_UB(res0, res1, dst, dst_stride);
+        ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -716,7 +716,7 @@  static void horiz_mc_qpel_no_rnd_aver_src1_8width_msa(const uint8_t *src,
         inp2 = (v16u8) __msa_insve_d((v2i64) inp2, 1, (v2i64) inp3);
         res0 = __msa_ave_u_b(inp0, res0);
         res1 = __msa_ave_u_b(inp2, res1);
-        ST8x4_UB(res0, res1, dst, dst_stride);
+        ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -799,7 +799,7 @@  static void horiz_mc_qpel_avg_dst_aver_src0_8width_msa(const uint8_t *src,
         dst2 = (v16u8) __msa_insve_d((v2i64) dst2, 1, (v2i64) dst3);
         AVER_UB2_UB(inp0, res0, inp2, res1, res0, res1);
         AVER_UB2_UB(dst0, res0, dst2, res1, res0, res1);
-        ST8x4_UB(res0, res1, dst, dst_stride);
+        ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -876,7 +876,7 @@  static void horiz_mc_qpel_avg_dst_8width_msa(const uint8_t *src,
         dst0 = (v16u8) __msa_insve_d((v2i64) dst0, 1, (v2i64) dst1);
         dst2 = (v16u8) __msa_insve_d((v2i64) dst2, 1, (v2i64) dst3);
         AVER_UB2_UB(dst0, res0, dst2, res1, res0, res1);
-        ST8x4_UB(res0, res1, dst, dst_stride);
+        ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -956,7 +956,7 @@  static void horiz_mc_qpel_avg_dst_aver_src1_8width_msa(const uint8_t *src,
         dst2 = (v16u8) __msa_insve_d((v2i64) dst2, 1, (v2i64) dst3);
         AVER_UB2_UB(inp0, res0, inp2, res1, res0, res1);
         AVER_UB2_UB(dst0, res0, dst2, res1, res0, res1);
-        ST8x4_UB(res0, res1, dst, dst_stride);
+        ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -1031,8 +1031,7 @@  static void vert_mc_qpel_aver_src0_8x8_msa(const uint8_t *src,
     tmp0 = (v16u8) __msa_insve_d((v2i64) inp0, 1, (v2i64) inp1);
     tmp1 = (v16u8) __msa_insve_d((v2i64) inp2, 1, (v2i64) inp3);
     AVER_UB2_UB(res0, tmp0, res1, tmp1, res0, res1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 
     inp8 = LD_UB(src);
     res0 = APPLY_VERT_QPEL_FILTER_8BYTE(inp4, inp3, inp2, inp1,
@@ -1048,8 +1047,7 @@  static void vert_mc_qpel_aver_src0_8x8_msa(const uint8_t *src,
     tmp0 = (v16u8) __msa_insve_d((v2i64) inp4, 1, (v2i64) inp5);
     tmp1 = (v16u8) __msa_insve_d((v2i64) inp6, 1, (v2i64) inp7);
     AVER_UB2_UB(res0, tmp0, res1, tmp1, res0, res1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void vert_mc_qpel_aver_src0_16x16_msa(const uint8_t *src,
@@ -1220,8 +1218,7 @@  static void vert_mc_qpel_8x8_msa(const uint8_t *src,
                                         inp3, inp2, inp1, inp0,
                                         inp4, inp5, inp6, inp7,
                                         const20, const6, const3);
-    ST8x4_UB(res0, res1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 
     inp8 = LD_UB(src);
     res0 = APPLY_VERT_QPEL_FILTER_8BYTE(inp4, inp3, inp2, inp1,
@@ -1234,8 +1231,7 @@  static void vert_mc_qpel_8x8_msa(const uint8_t *src,
                                         inp7, inp6, inp5, inp4,
                                         inp8, inp8, inp7, inp6,
                                         const20, const6, const3);
-    ST8x4_UB(res0, res1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void vert_mc_qpel_16x16_msa(const uint8_t *src,
@@ -1405,8 +1401,7 @@  static void vert_mc_qpel_aver_src1_8x8_msa(const uint8_t *src,
     tmp0 = (v16u8) __msa_insve_d((v2i64) inp1, 1, (v2i64) inp2);
     tmp1 = (v16u8) __msa_insve_d((v2i64) inp3, 1, (v2i64) inp4);
     AVER_UB2_UB(res0, tmp0, res1, tmp1, res0, res1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 
     inp8 = LD_UB(src);
     res0 = APPLY_VERT_QPEL_FILTER_8BYTE(inp4, inp3, inp2, inp1,
@@ -1422,7 +1417,7 @@  static void vert_mc_qpel_aver_src1_8x8_msa(const uint8_t *src,
     tmp0 = (v16u8) __msa_insve_d((v2i64) inp5, 1, (v2i64) inp6);
     tmp1 = (v16u8) __msa_insve_d((v2i64) inp7, 1, (v2i64) inp8);
     AVER_UB2_UB(res0, tmp0, res1, tmp1, res0, res1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void vert_mc_qpel_aver_src1_16x16_msa(const uint8_t *src,
@@ -1607,8 +1602,7 @@  static void vert_mc_qpel_no_rnd_aver_src0_8x8_msa(const uint8_t *src,
     tmp1 = (v16u8) __msa_insve_d((v2i64) inp2, 1, (v2i64) inp3);
     res0 = __msa_ave_u_b(res0, tmp0);
     res1 = __msa_ave_u_b(res1, tmp1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 
     inp8 = LD_UB(src);
     res0 = APPLY_VERT_QPEL_NO_ROUND_FILTER_8BYTE(inp4, inp3, inp2, inp1,
@@ -1625,8 +1619,7 @@  static void vert_mc_qpel_no_rnd_aver_src0_8x8_msa(const uint8_t *src,
     tmp1 = (v16u8) __msa_insve_d((v2i64) inp6, 1, (v2i64) inp7);
     res0 = __msa_ave_u_b(res0, tmp0);
     res1 = __msa_ave_u_b(res1, tmp1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void vert_mc_qpel_no_rnd_aver_src0_16x16_msa(const uint8_t *src,
@@ -1806,8 +1799,7 @@  static void vert_mc_qpel_no_rnd_8x8_msa(const uint8_t *src,
                                                  inp3, inp2, inp1, inp0,
                                                  inp4, inp5, inp6, inp7,
                                                  const20, const6, const3);
-    ST8x4_UB(res0, res1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 
     inp8 = LD_UB(src);
     res0 = APPLY_VERT_QPEL_NO_ROUND_FILTER_8BYTE(inp4, inp3, inp2, inp1,
@@ -1820,8 +1812,7 @@  static void vert_mc_qpel_no_rnd_8x8_msa(const uint8_t *src,
                                                  inp7, inp6, inp5, inp4,
                                                  inp8, inp8, inp7, inp6,
                                                  const20, const6, const3);
-    ST8x4_UB(res0, res1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void vert_mc_qpel_no_rnd_16x16_msa(const uint8_t *src,
@@ -1988,8 +1979,7 @@  static void vert_mc_qpel_no_rnd_aver_src1_8x8_msa(const uint8_t *src,
     tmp1 = (v16u8) __msa_insve_d((v2i64) inp3, 1, (v2i64) inp4);
     res0 = __msa_ave_u_b(res0, tmp0);
     res1 = __msa_ave_u_b(res1, tmp1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 
     inp8 = LD_UB(src);
     res0 = APPLY_VERT_QPEL_NO_ROUND_FILTER_8BYTE(inp4, inp3, inp2, inp1,
@@ -2006,7 +1996,7 @@  static void vert_mc_qpel_no_rnd_aver_src1_8x8_msa(const uint8_t *src,
     tmp1 = (v16u8) __msa_insve_d((v2i64) inp7, 1, (v2i64) inp8);
     res0 = __msa_ave_u_b(res0, tmp0);
     res1 = __msa_ave_u_b(res1, tmp1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
 }
 
 static void vert_mc_qpel_no_rnd_aver_src1_16x16_msa(const uint8_t *src,
@@ -2195,7 +2185,7 @@  static void vert_mc_qpel_avg_dst_aver_src0_8x8_msa(const uint8_t *src,
     dst2 = (v16u8) __msa_insve_d((v2i64) dst2, 1, (v2i64) dst3);
     AVER_UB2_UB(res0, tmp0, res1, tmp1, res0, res1);
     AVER_UB2_UB(dst0, res0, dst2, res1, res0, res1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
     dst += (4 * dst_stride);
 
     inp8 = LD_UB(src);
@@ -2217,7 +2207,7 @@  static void vert_mc_qpel_avg_dst_aver_src0_8x8_msa(const uint8_t *src,
     dst2 = (v16u8) __msa_insve_d((v2i64) dst2, 1, (v2i64) dst3);
     AVER_UB2_UB(res0, tmp0, res1, tmp1, res0, res1);
     AVER_UB2_UB(dst0, res0, dst2, res1, res0, res1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void vert_mc_qpel_avg_dst_aver_src0_16x16_msa(const uint8_t *src,
@@ -2384,7 +2374,7 @@  static void vert_mc_qpel_avg_dst_8x8_msa(const uint8_t *src,
     dst0 = (v16u8) __msa_insve_d((v2i64) dst0, 1, (v2i64) dst1);
     dst2 = (v16u8) __msa_insve_d((v2i64) dst2, 1, (v2i64) dst3);
     AVER_UB2_UB(dst0, res0, dst2, res1, res0, res1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
     dst += (4 * dst_stride);
 
     inp8 = LD_UB(src);
@@ -2402,8 +2392,7 @@  static void vert_mc_qpel_avg_dst_8x8_msa(const uint8_t *src,
     dst0 = (v16u8) __msa_insve_d((v2i64) dst0, 1, (v2i64) dst1);
     dst2 = (v16u8) __msa_insve_d((v2i64) dst2, 1, (v2i64) dst3);
     AVER_UB2_UB(dst0, res0, dst2, res1, res0, res1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void vert_mc_qpel_avg_dst_16x16_msa(const uint8_t *src,
@@ -2566,7 +2555,7 @@  static void vert_mc_qpel_avg_dst_aver_src1_8x8_msa(const uint8_t *src,
     dst2 = (v16u8) __msa_insve_d((v2i64) dst2, 1, (v2i64) dst3);
     AVER_UB2_UB(res0, tmp0, res1, tmp1, res0, res1);
     AVER_UB2_UB(dst0, res0, dst2, res1, res0, res1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
     dst += (4 * dst_stride);
 
     inp8 = LD_UB(src);
@@ -2587,7 +2576,7 @@  static void vert_mc_qpel_avg_dst_aver_src1_8x8_msa(const uint8_t *src,
     dst2 = (v16u8) __msa_insve_d((v2i64) dst2, 1, (v2i64) dst3);
     AVER_UB2_UB(res0, tmp0, res1, tmp1, res0, res1);
     AVER_UB2_UB(dst0, res0, dst2, res1, res0, res1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void vert_mc_qpel_avg_dst_aver_src1_16x16_msa(const uint8_t *src,
@@ -2832,7 +2821,7 @@  static void hv_mc_qpel_no_rnd_aver_hv_src00_8x8_msa(const uint8_t *src,
                                                  const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz1, (v2i64) horiz0);
     res0 = __msa_ave_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -2860,7 +2849,7 @@  static void hv_mc_qpel_no_rnd_aver_hv_src00_8x8_msa(const uint8_t *src,
                                                  horiz5, horiz4, horiz3, horiz2,
                                                  horiz6, horiz7, horiz8, horiz8,
                                                  const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz5, (v2i64) horiz4);
@@ -2870,12 +2859,12 @@  static void hv_mc_qpel_no_rnd_aver_hv_src00_8x8_msa(const uint8_t *src,
                                                  horiz7, horiz6, horiz5, horiz4,
                                                  horiz8, horiz8, horiz7, horiz6,
                                                  const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz7, (v2i64) horiz6);
     res1 = __msa_ave_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_no_rnd_horiz_16x16_msa(const uint8_t *src,
@@ -2977,7 +2966,7 @@  static void hv_mc_qpel_no_rnd_aver_v_src0_8x8_msa(const uint8_t *src,
                                                  const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz1, (v2i64) horiz0);
     res0 = __msa_ave_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -2999,7 +2988,7 @@  static void hv_mc_qpel_no_rnd_aver_v_src0_8x8_msa(const uint8_t *src,
     res1 = __msa_ave_u_b(avg1, res1);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz1, (v2i64) horiz0);
     res0 = __msa_ave_u_b(avg0, res0);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     res0 = APPLY_VERT_QPEL_NO_ROUND_FILTER_8BYTE(horiz4, horiz3, horiz2, horiz1,
@@ -3009,7 +2998,7 @@  static void hv_mc_qpel_no_rnd_aver_v_src0_8x8_msa(const uint8_t *src,
                                                  const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz5, (v2i64) horiz4);
     res0 = __msa_ave_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     res1 = APPLY_VERT_QPEL_NO_ROUND_FILTER_8BYTE(horiz6, horiz5, horiz4, horiz3,
@@ -3019,7 +3008,7 @@  static void hv_mc_qpel_no_rnd_aver_v_src0_8x8_msa(const uint8_t *src,
                                                  const20, const6, const3);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz7, (v2i64) horiz6);
     res1 = __msa_ave_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_no_rnd_horiz_src1_16x16_msa(const uint8_t *src,
@@ -3137,7 +3126,7 @@  static void hv_mc_qpel_no_rnd_aver_hv_src10_8x8_msa(const uint8_t *src,
                                                  const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz1, (v2i64) horiz0);
     res0 = __msa_ave_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -3163,7 +3152,7 @@  static void hv_mc_qpel_no_rnd_aver_hv_src10_8x8_msa(const uint8_t *src,
                                                  const20, const6, const3);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz3, (v2i64) horiz2);
     res1 = __msa_ave_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     res0 = APPLY_VERT_QPEL_NO_ROUND_FILTER_8BYTE(horiz4, horiz3, horiz2, horiz1,
@@ -3173,7 +3162,7 @@  static void hv_mc_qpel_no_rnd_aver_hv_src10_8x8_msa(const uint8_t *src,
                                                  const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz5, (v2i64) horiz4);
     res0 = __msa_ave_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     res1 = APPLY_VERT_QPEL_NO_ROUND_FILTER_8BYTE(horiz6, horiz5, horiz4, horiz3,
@@ -3183,7 +3172,7 @@  static void hv_mc_qpel_no_rnd_aver_hv_src10_8x8_msa(const uint8_t *src,
                                                  const20, const6, const3);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz7, (v2i64) horiz6);
     res1 = __msa_ave_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_no_rnd_aver_h_src0_16x16_msa(const uint8_t *src,
@@ -3246,7 +3235,7 @@  static void hv_mc_qpel_no_rnd_aver_h_src0_8x8_msa(const uint8_t *src,
 
     LD_UB2(src, src_stride, inp2, inp3);
     src += (2 * src_stride);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     res1 = APPLY_HORIZ_QPEL_NO_ROUND_FILTER_8BYTE(inp2, inp3, mask0, mask1,
@@ -3270,18 +3259,15 @@  static void hv_mc_qpel_no_rnd_aver_h_src0_8x8_msa(const uint8_t *src,
                                                  horiz5, horiz4, horiz3, horiz2,
                                                  horiz6, horiz7, horiz8, horiz8,
                                                  const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
-    dst += 2 * dst_stride;
-
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += (2 * dst_stride);
+    ST_D4(res1, res0, 0, 1, 0, 1, dst, dst_stride);
+    dst += (4 * dst_stride);
 
     res1 = APPLY_VERT_QPEL_NO_ROUND_FILTER_8BYTE(horiz6, horiz5, horiz4, horiz3,
                                                  horiz7, horiz8, horiz8, horiz7,
                                                  horiz7, horiz6, horiz5, horiz4,
                                                  horiz8, horiz8, horiz7, horiz6,
                                                  const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_no_rnd_16x16_msa(const uint8_t *src,
@@ -3337,7 +3323,7 @@  static void hv_mc_qpel_no_rnd_8x8_msa(const uint8_t *src,
                                                  const20, const6, const3);
     LD_UB2(src, src_stride, inp2, inp3);
     src += (2 * src_stride);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     horiz6 = APPLY_HORIZ_QPEL_NO_ROUND_FILTER_8BYTE(inp2, inp3, mask0, mask1,
@@ -3358,7 +3344,7 @@  static void hv_mc_qpel_no_rnd_8x8_msa(const uint8_t *src,
                                                  horiz5, horiz4, horiz3, horiz2,
                                                  horiz6, horiz7, horiz8, horiz8,
                                                  const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
 
@@ -3367,9 +3353,7 @@  static void hv_mc_qpel_no_rnd_8x8_msa(const uint8_t *src,
                                                  horiz7, horiz6, horiz5, horiz4,
                                                  horiz8, horiz8, horiz7, horiz6,
                                                  const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += 2 * dst_stride;
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_no_rnd_aver_h_src1_16x16_msa(const uint8_t *src,
@@ -3437,7 +3421,7 @@  static void hv_mc_qpel_no_rnd_aver_h_src1_8x8_msa(const uint8_t *src,
                                                  const20, const6, const3);
     LD_UB2(src, src_stride, inp2, inp3);
     src += (2 * src_stride);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     res1 = APPLY_HORIZ_QPEL_NO_ROUND_FILTER_8BYTE(inp2, inp3, mask0, mask1,
@@ -3464,7 +3448,7 @@  static void hv_mc_qpel_no_rnd_aver_h_src1_8x8_msa(const uint8_t *src,
                                                  horiz5, horiz4, horiz3, horiz2,
                                                  horiz6, horiz7, horiz8, horiz8,
                                                  const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     res1 = APPLY_VERT_QPEL_NO_ROUND_FILTER_8BYTE(horiz6, horiz5, horiz4, horiz3,
@@ -3472,9 +3456,7 @@  static void hv_mc_qpel_no_rnd_aver_h_src1_8x8_msa(const uint8_t *src,
                                                  horiz7, horiz6, horiz5, horiz4,
                                                  horiz8, horiz8, horiz7, horiz6,
                                                  const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += 2 * dst_stride;
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_no_rnd_aver_hv_src01_16x16_msa(const uint8_t *src,
@@ -3536,7 +3518,7 @@  static void hv_mc_qpel_no_rnd_aver_hv_src01_8x8_msa(const uint8_t *src,
                                                  const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz2, (v2i64) horiz1);
     res0 = __msa_ave_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -3564,7 +3546,7 @@  static void hv_mc_qpel_no_rnd_aver_hv_src01_8x8_msa(const uint8_t *src,
                                                  horiz5, horiz4, horiz3, horiz2,
                                                  horiz6, horiz7, horiz8, horiz8,
                                                  const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz6, (v2i64) horiz5);
@@ -3575,12 +3557,12 @@  static void hv_mc_qpel_no_rnd_aver_hv_src01_8x8_msa(const uint8_t *src,
                                                  horiz7, horiz6, horiz5, horiz4,
                                                  horiz8, horiz8, horiz7, horiz6,
                                                  const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz8, (v2i64) horiz7);
     res1 = __msa_ave_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_no_rnd_aver_v_src1_16x16_msa(const uint8_t *src,
@@ -3638,7 +3620,7 @@  static void hv_mc_qpel_no_rnd_aver_v_src1_8x8_msa(const uint8_t *src,
     res0 = __msa_ave_u_b(avg0, res0);
     LD_UB2(src, src_stride, inp2, inp3);
     src += (2 * src_stride);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     horiz6 = APPLY_HORIZ_QPEL_NO_ROUND_FILTER_8BYTE(inp2, inp3, mask0, mask1,
@@ -3656,7 +3638,7 @@  static void hv_mc_qpel_no_rnd_aver_v_src1_8x8_msa(const uint8_t *src,
     horiz8 = APPLY_HORIZ_QPEL_NO_ROUND_FILTER_8BYTE_1ROW(inp0, mask0, mask1,
                                                          mask2, mask3, const20,
                                                          const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     res0 = APPLY_VERT_QPEL_NO_ROUND_FILTER_8BYTE(horiz4, horiz3, horiz2, horiz1,
@@ -3671,12 +3653,9 @@  static void hv_mc_qpel_no_rnd_aver_v_src1_8x8_msa(const uint8_t *src,
                                                  horiz7, horiz6, horiz5, horiz4,
                                                  horiz8, horiz8, horiz7, horiz6,
                                                  const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += 2 * dst_stride;
-
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz8, (v2i64) horiz7);
     res1 = __msa_ave_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_no_rnd_aver_hv_src11_16x16_msa(const uint8_t *src,
@@ -3744,7 +3723,7 @@  static void hv_mc_qpel_no_rnd_aver_hv_src11_8x8_msa(const uint8_t *src,
                                                  const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz2, (v2i64) horiz1);
     res0 = __msa_ave_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -3764,7 +3743,7 @@  static void hv_mc_qpel_no_rnd_aver_hv_src11_8x8_msa(const uint8_t *src,
                                                  const20, const6, const3);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz4, (v2i64) horiz3);
     res1 = __msa_ave_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     inp0 = LD_UB(src);
@@ -3787,7 +3766,7 @@  static void hv_mc_qpel_no_rnd_aver_hv_src11_8x8_msa(const uint8_t *src,
     res0 = __msa_ave_u_b(avg0, res0);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz8, (v2i64) horiz7);
     res1 = __msa_ave_u_b(avg1, res1);
-    ST8x4_UB(res0, res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_aver_horiz_src0_16x16_msa(const uint8_t *src,
@@ -3893,7 +3872,7 @@  static void hv_mc_qpel_aver_hv_src00_8x8_msa(const uint8_t *src,
                                         const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz1, (v2i64) horiz0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -3915,7 +3894,7 @@  static void hv_mc_qpel_aver_hv_src00_8x8_msa(const uint8_t *src,
     res0 = APPLY_HORIZ_QPEL_FILTER_8BYTE_1ROW(inp0, mask0, mask1, mask2, mask3,
                                               const20, const6, const3);
     horiz8 = __msa_aver_u_b(inp0, res0);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     res0 = APPLY_VERT_QPEL_FILTER_8BYTE(horiz4, horiz3, horiz2, horiz1,
@@ -3930,11 +3909,9 @@  static void hv_mc_qpel_aver_hv_src00_8x8_msa(const uint8_t *src,
                                         horiz7, horiz6, horiz5, horiz4,
                                         horiz8, horiz8, horiz7, horiz6,
                                         const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += 2 * dst_stride;
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz7, (v2i64) horiz6);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_aver_horiz_16x16_msa(const uint8_t *src,
@@ -4034,7 +4011,7 @@  static void hv_mc_qpel_aver_v_src0_8x8_msa(const uint8_t *src,
                                         const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz1, (v2i64) horiz0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -4059,7 +4036,7 @@  static void hv_mc_qpel_aver_v_src0_8x8_msa(const uint8_t *src,
                                         horiz5, horiz4, horiz3, horiz2,
                                         horiz6, horiz7, horiz8, horiz8,
                                         const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz5, (v2i64) horiz4);
@@ -4069,11 +4046,9 @@  static void hv_mc_qpel_aver_v_src0_8x8_msa(const uint8_t *src,
                                         horiz7, horiz6, horiz5, horiz4,
                                         horiz8, horiz8, horiz7, horiz6,
                                         const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += 2 * dst_stride;
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz7, (v2i64) horiz6);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_aver_horiz_src1_16x16_msa(const uint8_t *src,
@@ -4197,7 +4172,7 @@  static void hv_mc_qpel_aver_hv_src10_8x8_msa(const uint8_t *src,
                                         horiz3, horiz2, horiz1, horiz0,
                                         horiz4, horiz5, horiz6, horiz7,
                                         const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     inp0 = LD_UB(src);
@@ -4212,7 +4187,7 @@  static void hv_mc_qpel_aver_hv_src10_8x8_msa(const uint8_t *src,
                                         horiz5, horiz4, horiz3, horiz2,
                                         horiz6, horiz7, horiz8, horiz8,
                                         const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz5, (v2i64) horiz4);
@@ -4222,12 +4197,9 @@  static void hv_mc_qpel_aver_hv_src10_8x8_msa(const uint8_t *src,
                                         horiz7, horiz6, horiz5, horiz4,
                                         horiz8, horiz8, horiz7, horiz6,
                                         const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += 2 * dst_stride;
-
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz7, (v2i64) horiz6);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_aver_h_src0_16x16_msa(const uint8_t *src,
@@ -4285,7 +4257,7 @@  static void hv_mc_qpel_aver_h_src0_8x8_msa(const uint8_t *src,
                                         horiz1, horiz0, horiz0, horiz1,
                                         horiz2, horiz3, horiz4, horiz5,
                                         const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -4309,7 +4281,7 @@  static void hv_mc_qpel_aver_h_src0_8x8_msa(const uint8_t *src,
                                         horiz5, horiz4, horiz3, horiz2,
                                         horiz6, horiz7, horiz8, horiz8,
                                         const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     res1 = APPLY_VERT_QPEL_FILTER_8BYTE(horiz6, horiz5, horiz4, horiz3,
@@ -4317,9 +4289,7 @@  static void hv_mc_qpel_aver_h_src0_8x8_msa(const uint8_t *src,
                                         horiz7, horiz6, horiz5, horiz4,
                                         horiz8, horiz8, horiz7, horiz6,
                                         const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += 2 * dst_stride;
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_16x16_msa(const uint8_t *src,
@@ -4371,7 +4341,7 @@  static void hv_mc_qpel_8x8_msa(const uint8_t *src, int32_t src_stride,
                                         horiz1, horiz0, horiz0, horiz1,
                                         horiz2, horiz3, horiz4, horiz5,
                                         const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -4389,7 +4359,7 @@  static void hv_mc_qpel_8x8_msa(const uint8_t *src, int32_t src_stride,
     horiz8 = APPLY_HORIZ_QPEL_FILTER_8BYTE_1ROW(inp0,
                                                 mask0, mask1, mask2, mask3,
                                                 const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     res0 = APPLY_VERT_QPEL_FILTER_8BYTE(horiz4, horiz3, horiz2, horiz1,
@@ -4402,9 +4372,7 @@  static void hv_mc_qpel_8x8_msa(const uint8_t *src, int32_t src_stride,
                                         horiz7, horiz6, horiz5, horiz4,
                                         horiz8, horiz8, horiz7, horiz6,
                                         const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += 2 * dst_stride;
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_aver_h_src1_16x16_msa(const uint8_t *src,
@@ -4478,31 +4446,25 @@  static void hv_mc_qpel_aver_h_src1_8x8_msa(const uint8_t *src,
                                         horiz1, horiz0, horiz0, horiz1,
                                         horiz2, horiz3, horiz4, horiz5,
                                         const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += (2 * dst_stride);
-
     res1 = APPLY_VERT_QPEL_FILTER_8BYTE(horiz2, horiz1, horiz0, horiz0,
                                         horiz3, horiz4, horiz5, horiz6,
                                         horiz3, horiz2, horiz1, horiz0,
                                         horiz4, horiz5, horiz6, horiz7,
                                         const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
-    dst += (2 * dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
+    dst += (4 * dst_stride);
 
     res0 = APPLY_VERT_QPEL_FILTER_8BYTE(horiz4, horiz3, horiz2, horiz1,
                                         horiz5, horiz6, horiz7, horiz8,
                                         horiz5, horiz4, horiz3, horiz2,
                                         horiz6, horiz7, horiz8, horiz8,
                                         const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += (2 * dst_stride);
-
     res1 = APPLY_VERT_QPEL_FILTER_8BYTE(horiz6, horiz5, horiz4, horiz3,
                                         horiz7, horiz8, horiz8, horiz7,
                                         horiz7, horiz6, horiz5, horiz4,
                                         horiz8, horiz8, horiz7, horiz6,
                                         const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_aver_hv_src01_16x16_msa(const uint8_t *src,
@@ -4561,7 +4523,7 @@  static void hv_mc_qpel_aver_hv_src01_8x8_msa(const uint8_t *src,
                                         const20, const6, const3);
     avg0 = (v16u8) __msa_insve_d((v2i64) horiz1, 1, (v2i64) horiz2);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -4587,7 +4549,7 @@  static void hv_mc_qpel_aver_hv_src01_8x8_msa(const uint8_t *src,
                                         horiz5, horiz4, horiz3, horiz2,
                                         horiz6, horiz7, horiz8, horiz8,
                                         const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     avg0 = (v16u8) __msa_insve_d((v2i64) horiz5, 1, (v2i64) horiz6);
@@ -4597,13 +4559,9 @@  static void hv_mc_qpel_aver_hv_src01_8x8_msa(const uint8_t *src,
                                         horiz7, horiz6, horiz5, horiz4,
                                         horiz8, horiz8, horiz7, horiz6,
                                         const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += 2 * dst_stride;
-
     avg1 = (v16u8) __msa_insve_d((v2i64) horiz7, 1, (v2i64) horiz8);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
-    dst += (2 * dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_aver_v_src1_16x16_msa(const uint8_t *src,
@@ -4660,7 +4618,7 @@  static void hv_mc_qpel_aver_v_src1_8x8_msa(const uint8_t *src,
                                         const20, const6, const3);
     avg0 = (v16u8) __msa_insve_d((v2i64) horiz1, 1, (v2i64) horiz2);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -4685,7 +4643,7 @@  static void hv_mc_qpel_aver_v_src1_8x8_msa(const uint8_t *src,
                                         horiz5, horiz4, horiz3, horiz2,
                                         horiz6, horiz7, horiz8, horiz8,
                                         const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
     avg0 = (v16u8) __msa_insve_d((v2i64) horiz5, 1, (v2i64) horiz6);
     res0 = __msa_aver_u_b(avg0, res0);
@@ -4695,11 +4653,9 @@  static void hv_mc_qpel_aver_v_src1_8x8_msa(const uint8_t *src,
                                         horiz7, horiz6, horiz5, horiz4,
                                         horiz8, horiz8, horiz7, horiz6,
                                         const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += 2 * dst_stride;
     avg1 = (v16u8) __msa_insve_d((v2i64) horiz7, 1, (v2i64) horiz8);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_aver_hv_src11_16x16_msa(const uint8_t *src,
@@ -4764,7 +4720,7 @@  static void hv_mc_qpel_aver_hv_src11_8x8_msa(const uint8_t *src,
     res0 = __msa_aver_u_b(avg0, res0);
     LD_UB2(src, src_stride, inp2, inp3);
     src += (2 * src_stride);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     res1 = APPLY_HORIZ_QPEL_FILTER_8BYTE(inp2, inp3, mask0, mask1, mask2, mask3,
@@ -4791,7 +4747,7 @@  static void hv_mc_qpel_aver_hv_src11_8x8_msa(const uint8_t *src,
                                         horiz5, horiz4, horiz3, horiz2,
                                         horiz6, horiz7, horiz8, horiz8,
                                         const20, const6, const3);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += 2 * dst_stride;
 
     avg0 = (v16u8) __msa_ilvr_d((v2i64) horiz6, (v2i64) horiz5);
@@ -4801,12 +4757,9 @@  static void hv_mc_qpel_aver_hv_src11_8x8_msa(const uint8_t *src,
                                         horiz7, horiz6, horiz5, horiz4,
                                         horiz8, horiz8, horiz7, horiz6,
                                         const20, const6, const3);
-    ST8x2_UB(res0, dst, dst_stride);
-    dst += 2 * dst_stride;
-
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz8, (v2i64) horiz7);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D4(res0, res1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_avg_dst_aver_hv_src00_16x16_msa(const uint8_t *src,
@@ -4869,7 +4822,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src00_8x8_msa(const uint8_t *src,
     res0 = __msa_aver_u_b(avg0, res0);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -4889,7 +4842,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src00_8x8_msa(const uint8_t *src,
     res1 = __msa_aver_u_b(avg1, res1);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     inp0 = LD_UB(src);
@@ -4906,7 +4859,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src00_8x8_msa(const uint8_t *src,
     res0 = __msa_aver_u_b(avg0, res0);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(dst, dst_stride, dst0, dst1);
@@ -4919,7 +4872,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src00_8x8_msa(const uint8_t *src,
     res1 = __msa_aver_u_b(avg1, res1);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_avg_dst_aver_v_src0_16x16_msa(const uint8_t *src,
@@ -4979,7 +4932,7 @@  static void hv_mc_qpel_avg_dst_aver_v_src0_8x8_msa(const uint8_t *src,
     res0 = __msa_aver_u_b(avg0, res0);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -4998,7 +4951,7 @@  static void hv_mc_qpel_avg_dst_aver_v_src0_8x8_msa(const uint8_t *src,
     res1 = __msa_aver_u_b(avg1, res1);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     inp0 = LD_UB(src);
@@ -5015,7 +4968,7 @@  static void hv_mc_qpel_avg_dst_aver_v_src0_8x8_msa(const uint8_t *src,
     res0 = __msa_aver_u_b(avg0, res0);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(dst, dst_stride, dst0, dst1);
@@ -5028,7 +4981,7 @@  static void hv_mc_qpel_avg_dst_aver_v_src0_8x8_msa(const uint8_t *src,
     res1 = __msa_aver_u_b(avg1, res1);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_avg_dst_aver_hv_src10_16x16_msa(const uint8_t *src,
@@ -5099,7 +5052,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src10_8x8_msa(const uint8_t *src,
     res0 = __msa_aver_u_b(avg0, res0);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -5122,7 +5075,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src10_8x8_msa(const uint8_t *src,
     res1 = __msa_aver_u_b(avg1, res1);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     inp0 = LD_UB(src);
@@ -5140,7 +5093,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src10_8x8_msa(const uint8_t *src,
     res0 = __msa_aver_u_b(avg0, res0);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(dst, dst_stride, dst0, dst1);
@@ -5153,7 +5106,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src10_8x8_msa(const uint8_t *src,
     res1 = __msa_aver_u_b(avg1, res1);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_avg_dst_aver_h_src0_16x16_msa(const uint8_t *src,
@@ -5214,7 +5167,7 @@  static void hv_mc_qpel_avg_dst_aver_h_src0_8x8_msa(const uint8_t *src,
                                         const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -5232,7 +5185,7 @@  static void hv_mc_qpel_avg_dst_aver_h_src0_8x8_msa(const uint8_t *src,
                                         const20, const6, const3);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     inp0 = LD_UB(src);
@@ -5247,7 +5200,7 @@  static void hv_mc_qpel_avg_dst_aver_h_src0_8x8_msa(const uint8_t *src,
                                         const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(dst, dst_stride, dst0, dst1);
@@ -5258,8 +5211,7 @@  static void hv_mc_qpel_avg_dst_aver_h_src0_8x8_msa(const uint8_t *src,
                                         const20, const6, const3);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
-    dst += (2 * dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_avg_dst_16x16_msa(const uint8_t *src, int32_t src_stride,
@@ -5324,7 +5276,7 @@  static void hv_mc_qpel_avg_dst_8x8_msa(const uint8_t *src, int32_t src_stride,
                                         const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(dst, dst_stride, dst0, dst1);
@@ -5335,7 +5287,7 @@  static void hv_mc_qpel_avg_dst_8x8_msa(const uint8_t *src, int32_t src_stride,
                                         const20, const6, const3);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(dst, dst_stride, dst0, dst1);
@@ -5346,7 +5298,7 @@  static void hv_mc_qpel_avg_dst_8x8_msa(const uint8_t *src, int32_t src_stride,
                                         const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(dst, dst_stride, dst0, dst1);
@@ -5357,7 +5309,7 @@  static void hv_mc_qpel_avg_dst_8x8_msa(const uint8_t *src, int32_t src_stride,
                                         const20, const6, const3);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_avg_dst_aver_h_src1_16x16_msa(const uint8_t *src,
@@ -5425,7 +5377,7 @@  static void hv_mc_qpel_avg_dst_aver_h_src1_8x8_msa(const uint8_t *src,
                                         const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -5446,7 +5398,7 @@  static void hv_mc_qpel_avg_dst_aver_h_src1_8x8_msa(const uint8_t *src,
                                         const20, const6, const3);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     inp0 = LD_UB(src);
@@ -5462,7 +5414,7 @@  static void hv_mc_qpel_avg_dst_aver_h_src1_8x8_msa(const uint8_t *src,
                                         const20, const6, const3);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(dst, dst_stride, dst0, dst1);
@@ -5473,7 +5425,7 @@  static void hv_mc_qpel_avg_dst_aver_h_src1_8x8_msa(const uint8_t *src,
                                         const20, const6, const3);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_avg_dst_aver_hv_src01_16x16_msa(const uint8_t *src,
@@ -5537,7 +5489,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src01_8x8_msa(const uint8_t *src,
     res0 = __msa_aver_u_b(avg0, res0);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(dst, dst_stride, dst0, dst1);
@@ -5557,7 +5509,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src01_8x8_msa(const uint8_t *src,
     res1 = __msa_aver_u_b(avg1, res1);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     inp0 = LD_UB(src);
@@ -5579,7 +5531,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src01_8x8_msa(const uint8_t *src,
     LD_UB2(dst, dst_stride, dst0, dst1);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz8, (v2i64) horiz7);
@@ -5587,7 +5539,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src01_8x8_msa(const uint8_t *src,
     LD_UB2(dst, dst_stride, dst0, dst1);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_avg_dst_aver_v_src1_16x16_msa(const uint8_t *src,
@@ -5647,7 +5599,7 @@  static void hv_mc_qpel_avg_dst_aver_v_src1_8x8_msa(const uint8_t *src,
     res0 = __msa_aver_u_b(avg0, res0);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(dst, dst_stride, dst0, dst1);
@@ -5666,7 +5618,7 @@  static void hv_mc_qpel_avg_dst_aver_v_src1_8x8_msa(const uint8_t *src,
     res1 = __msa_aver_u_b(avg1, res1);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     inp0 = LD_UB(src);
@@ -5686,7 +5638,7 @@  static void hv_mc_qpel_avg_dst_aver_v_src1_8x8_msa(const uint8_t *src,
     LD_UB2(dst, dst_stride, dst0, dst1);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     avg1 = (v16u8) __msa_ilvr_d((v2i64) horiz8, (v2i64) horiz7);
@@ -5694,7 +5646,7 @@  static void hv_mc_qpel_avg_dst_aver_v_src1_8x8_msa(const uint8_t *src,
     LD_UB2(dst, dst_stride, dst0, dst1);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void hv_mc_qpel_avg_dst_aver_hv_src11_16x16_msa(const uint8_t *src,
@@ -5762,7 +5714,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src11_8x8_msa(const uint8_t *src,
     res0 = __msa_aver_u_b(avg0, res0);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(src, src_stride, inp2, inp3);
@@ -5783,7 +5735,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src11_8x8_msa(const uint8_t *src,
     res1 = __msa_aver_u_b(avg1, res1);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     inp0 = LD_UB(src);
@@ -5800,7 +5752,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src11_8x8_msa(const uint8_t *src,
     res0 = __msa_aver_u_b(avg0, res0);
     avg0 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res0 = __msa_aver_u_b(avg0, res0);
-    ST8x2_UB(res0, dst, dst_stride);
+    ST_D2(res0, 0, 1, dst, dst_stride);
     dst += (2 * dst_stride);
 
     LD_UB2(dst, dst_stride, dst0, dst1);
@@ -5812,7 +5764,7 @@  static void hv_mc_qpel_avg_dst_aver_hv_src11_8x8_msa(const uint8_t *src,
     res1 = __msa_aver_u_b(avg1, res1);
     avg1 = (v16u8) __msa_ilvr_d((v2i64) dst1, (v2i64) dst0);
     res1 = __msa_aver_u_b(avg1, res1);
-    ST8x2_UB(res1, dst, dst_stride);
+    ST_D2(res1, 0, 1, dst, dst_stride);
 }
 
 static void copy_8x8_msa(const uint8_t *src, int32_t src_stride,
diff --git a/libavcodec/mips/vp3dsp_idct_msa.c b/libavcodec/mips/vp3dsp_idct_msa.c
index 5427ac5..b2899ee 100644
--- a/libavcodec/mips/vp3dsp_idct_msa.c
+++ b/libavcodec/mips/vp3dsp_idct_msa.c
@@ -47,7 +47,6 @@  static void idct_msa(uint8_t *dst, int stride, int16_t *input, int type)
     v4i32 cnst8w = {8, 8, 8, 8};
     v4i32 cnst2048w = {2048, 2048, 2048, 2048};
     v4i32 cnst128w = {128, 128, 128, 128};
-    int nstride = stride;
 
     /* Extended input data */
     LD_SH8(input, 8, r0, r1, r2, r3, r4, r5, r6, r7);
@@ -386,20 +385,14 @@  static void idct_msa(uint8_t *dst, int stride, int16_t *input, int type)
     VSHF_B2_SB(r2_l, r6_l, r3_l, r7_l, mask, mask, d6, d7);
 
     /* Final sequence of operations over-write original dst */
-    ST8x1_UB(d0, dst);
-    ST8x1_UB(d1, dst + nstride);
-    nstride += stride;
-    ST8x1_UB(d2, dst + nstride);
-    nstride += stride;
-    ST8x1_UB(d3, dst + nstride);
-    nstride += stride;
-    ST8x1_UB(d4, dst + nstride);
-    nstride += stride;
-    ST8x1_UB(d5, dst + nstride);
-    nstride += stride;
-    ST8x1_UB(d6, dst + nstride);
-    nstride += stride;
-    ST8x1_UB(d7, dst + nstride);
+    ST_D1(d0, 0, dst);
+    ST_D1(d1, 0, dst + stride);
+    ST_D1(d2, 0, dst + 2 * stride);
+    ST_D1(d3, 0, dst + 3 * stride);
+    ST_D1(d4, 0, dst + 4 * stride);
+    ST_D1(d5, 0, dst + 5 * stride);
+    ST_D1(d6, 0, dst + 6 * stride);
+    ST_D1(d7, 0, dst + 7 * stride);
 }
 
 void ff_vp3_idct_put_msa(uint8_t *dest, ptrdiff_t line_size, int16_t *block)
@@ -424,7 +417,6 @@  void ff_vp3_idct_dc_add_msa(uint8_t *dest, ptrdiff_t line_size, int16_t *block)
     v4i32 r0, r1, r2, r3, r4, r5, r6, r7;
     v16i8 mask = {0, 4, 8, 12, 16, 20, 24, 28, 0, 0, 0, 0, 0, 0, 0, 0};
     v16i8 zero = {0};
-    int nstride = line_size;
 
     LD_SB8(dest, line_size, d0, d1, d2, d3, d4, d5, d6, d7);
     ILVR_B4_SW(zero, d0, zero, d1, zero, d2, zero, d3,
@@ -480,20 +472,14 @@  void ff_vp3_idct_dc_add_msa(uint8_t *dest, ptrdiff_t line_size, int16_t *block)
     VSHF_B2_SB(e6, r6, e7, r7, mask, mask, d6, d7);
 
     /* Final sequence of operations over-write original dst */
-    ST8x1_UB(d0, dest);
-    ST8x1_UB(d1, dest + nstride);
-    nstride += line_size;
-    ST8x1_UB(d2, dest + nstride);
-    nstride += line_size;
-    ST8x1_UB(d3, dest + nstride);
-    nstride += line_size;
-    ST8x1_UB(d4, dest + nstride);
-    nstride += line_size;
-    ST8x1_UB(d5, dest + nstride);
-    nstride += line_size;
-    ST8x1_UB(d6, dest + nstride);
-    nstride += line_size;
-    ST8x1_UB(d7, dest + nstride);
+    ST_D1(d0, 0, dest);
+    ST_D1(d1, 0, dest + line_size);
+    ST_D1(d2, 0, dest + 2 * line_size);
+    ST_D1(d3, 0, dest + 3 * line_size);
+    ST_D1(d4, 0, dest + 4 * line_size);
+    ST_D1(d5, 0, dest + 5 * line_size);
+    ST_D1(d6, 0, dest + 6 * line_size);
+    ST_D1(d7, 0, dest + 7 * line_size);
 
     block[0] = 0;
 }
@@ -537,8 +523,8 @@  void ff_vp3_v_loop_filter_msa(uint8_t *first_pixel, ptrdiff_t stride,
     VSHF_B2_SB(f0, f1, g0, g1, mask, mask, d1, d2);
 
     /* Final move to first_pixel */
-    ST8x1_UB(d1, first_pixel + nstride);
-    ST8x1_UB(d2, first_pixel);
+    ST_D1(d1, 0, first_pixel + nstride);
+    ST_D1(d2, 0, first_pixel);
 }
 
 void ff_vp3_h_loop_filter_msa(uint8_t *first_pixel, ptrdiff_t stride,
@@ -583,8 +569,8 @@  void ff_vp3_h_loop_filter_msa(uint8_t *first_pixel, ptrdiff_t stride,
     g1 = CLIP_SW_0_255(g1);
     VSHF_B2_SB(f0, g0, f1, g1, mask, mask, d1, d2);
     /* Final move to first_pixel */
-    ST2x4_UB(d1, 0, first_pixel - 1, stride);
-    ST2x4_UB(d2, 0, first_pixel - 1 + 4 * stride, stride);
+    ST_H4(d1, 0, 1, 2, 3, first_pixel - 1, stride);
+    ST_H4(d2, 0, 1, 2, 3, first_pixel - 1 + 4 * stride, stride);
 }
 
 void ff_put_no_rnd_pixels_l2_msa(uint8_t *dst, const uint8_t *src1,
@@ -641,10 +627,8 @@  void ff_put_no_rnd_pixels_l2_msa(uint8_t *dst, const uint8_t *src1,
         f2 = (v4i32) __msa_and_v((v16u8)a3, (v16u8)b3);
         t3 = t3 + (v4u32)f2;
 
-        ST4x4_UB(t0, t0, 0, 1, 2, 3, dst, stride);
-        ST4x4_UB(t1, t1, 0, 1, 2, 3, dst + 4 * stride, stride);
-        ST4x4_UB(t2, t2, 0, 1, 2, 3, dst + 4, stride);
-        ST4x4_UB(t3, t3, 0, 1, 2, 3, dst + 4 + 4 * stride, stride);
+        ST_W8(t0, t1, 0, 1, 2, 3, 0, 1, 2, 3, dst, stride);
+        ST_W8(t2, t3, 0, 1, 2, 3, 0, 1, 2, 3, dst + 4, stride);
     } else {
         int i;
 
diff --git a/libavcodec/mips/vp8_idct_msa.c b/libavcodec/mips/vp8_idct_msa.c
index 11ac9ff..ae6fec0 100644
--- a/libavcodec/mips/vp8_idct_msa.c
+++ b/libavcodec/mips/vp8_idct_msa.c
@@ -76,7 +76,8 @@  void ff_vp8_idct_add_msa(uint8_t *dst, int16_t input[16], ptrdiff_t stride)
     res2 = CLIP_SW_0_255(res2);
     res3 = CLIP_SW_0_255(res3);
     VSHF_B2_SB(res0, res1, res2, res3, mask, mask, dest0, dest1);
-    ST4x4_UB(dest0, dest1, 0, 1, 0, 1, dst, stride);
+    ST_W2(dest0, 0, 1, dst, stride);
+    ST_W2(dest1, 0, 1, dst + 2 * stride, stride);
 
     memset(input, 0, 4 * 4 * sizeof(*input));
 }
@@ -97,7 +98,8 @@  void ff_vp8_idct_dc_add_msa(uint8_t *dst, int16_t in_dc[16], ptrdiff_t stride)
     ADD4(res0, vec, res1, vec, res2, vec, res3, vec, res0, res1, res2, res3);
     CLIP_SH4_0_255(res0, res1, res2, res3);
     VSHF_B2_SB(res0, res1, res2, res3, mask, mask, dest0, dest1);
-    ST4x4_UB(dest0, dest1, 0, 1, 0, 1, dst, stride);
+    ST_W2(dest0, 0, 1, dst, stride);
+    ST_W2(dest1, 0, 1, dst + 2 * stride, stride);
 
     in_dc[0] = 0;
 }
diff --git a/libavcodec/mips/vp8_lpf_msa.c b/libavcodec/mips/vp8_lpf_msa.c
index 3590961..1b51334 100644
--- a/libavcodec/mips/vp8_lpf_msa.c
+++ b/libavcodec/mips/vp8_lpf_msa.c
@@ -540,14 +540,8 @@  void ff_vp8_h_loop_filter_simple_msa(uint8_t *src, ptrdiff_t pitch,
     ILVRL_B2_SH(q0, p0, tmp1, tmp0);
 
     src -= 1;
-    ST2x4_UB(tmp1, 0, src, pitch);
-    src += 4 * pitch;
-    ST2x4_UB(tmp1, 4, src, pitch);
-    src += 4 * pitch;
-    ST2x4_UB(tmp0, 0, src, pitch);
-    src += 4 * pitch;
-    ST2x4_UB(tmp0, 4, src, pitch);
-    src += 4 * pitch;
+    ST_H8(tmp1, 0, 1, 2, 3, 4, 5, 6, 7, src, pitch)
+    ST_H8(tmp0, 0, 1, 2, 3, 4, 5, 6, 7, src + 8 * pitch, pitch)
 }
 
 void ff_vp8_v_loop_filter8uv_inner_msa(uint8_t *src_u, uint8_t *src_v,
@@ -596,7 +590,6 @@  void ff_vp8_h_loop_filter8uv_inner_msa(uint8_t *src_u, uint8_t *src_v,
                                        ptrdiff_t pitch, int b_limit_in,
                                        int limit_in, int thresh_in)
 {
-    uint8_t *temp_src_u, *temp_src_v;
     v16u8 p3, p2, p1, p0, q3, q2, q1, q0;
     v16u8 mask, hev, flat, thresh, limit, b_limit;
     v16u8 row0, row1, row2, row3, row4, row5, row6, row7, row8;
@@ -623,15 +616,8 @@  void ff_vp8_h_loop_filter8uv_inner_msa(uint8_t *src_u, uint8_t *src_v,
     tmp1 = (v4i32) __msa_ilvl_b((v16i8) q1, (v16i8) q0);
     ILVRL_H2_SW(tmp1, tmp0, tmp4, tmp5);
 
-    temp_src_u = src_u - 2;
-    ST4x4_UB(tmp2, tmp2, 0, 1, 2, 3, temp_src_u, pitch);
-    temp_src_u += 4 * pitch;
-    ST4x4_UB(tmp3, tmp3, 0, 1, 2, 3, temp_src_u, pitch);
-
-    temp_src_v = src_v - 2;
-    ST4x4_UB(tmp4, tmp4, 0, 1, 2, 3, temp_src_v, pitch);
-    temp_src_v += 4 * pitch;
-    ST4x4_UB(tmp5, tmp5, 0, 1, 2, 3, temp_src_v, pitch);
+    ST_W8(tmp2, tmp3, 0, 1, 2, 3, 0, 1, 2, 3, src_u - 2, pitch);
+    ST_W8(tmp4, tmp5, 0, 1, 2, 3, 0, 1, 2, 3, src_v - 2, pitch);
 }
 
 void ff_vp8_v_loop_filter16_inner_msa(uint8_t *src, ptrdiff_t pitch,
@@ -684,7 +670,6 @@  void ff_vp8_h_loop_filter16_inner_msa(uint8_t *src, ptrdiff_t pitch,
     ILVRL_H2_SH(tmp1, tmp0, tmp4, tmp5);
 
     src -= 2;
-    ST4x8_UB(tmp2, tmp3, src, pitch);
-    src += (8 * pitch);
-    ST4x8_UB(tmp4, tmp5, src, pitch);
+    ST_W8(tmp2, tmp3, 0, 1, 2, 3, 0, 1, 2, 3, src, pitch)
+    ST_W8(tmp4, tmp5, 0, 1, 2, 3, 0, 1, 2, 3, src + 8 * pitch, pitch)
 }
diff --git a/libavcodec/mips/vp8_mc_msa.c b/libavcodec/mips/vp8_mc_msa.c
index 2bf0abd..57af6b4 100644
--- a/libavcodec/mips/vp8_mc_msa.c
+++ b/libavcodec/mips/vp8_mc_msa.c
@@ -181,7 +181,7 @@  static void common_hz_6t_4x4_msa(uint8_t *src, int32_t src_stride,
     SRARI_H2_SH(out0, out1, 7);
     SAT_SH2_SH(out0, out1, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void common_hz_6t_4x8_msa(uint8_t *src, int32_t src_stride,
@@ -214,10 +214,9 @@  static void common_hz_6t_4x8_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_SH(out0, out1, out2, out3, 7);
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
     out = PCKEV_XORI128_UB(out2, out3);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
 }
 
 void ff_put_vp8_epel4_h6_msa(uint8_t *dst, ptrdiff_t dst_stride,
@@ -263,7 +262,7 @@  void ff_put_vp8_epel8_h6_msa(uint8_t *dst, ptrdiff_t dst_stride,
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     tmp0 = PCKEV_XORI128_UB(out0, out1);
     tmp1 = PCKEV_XORI128_UB(out2, out3);
-    ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+    ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
     dst += (4 * dst_stride);
 
     for (loop_cnt = (height >> 2) - 1; loop_cnt--;) {
@@ -276,7 +275,7 @@  void ff_put_vp8_epel8_h6_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SAT_SH4_SH(out0, out1, out2, out3, 7);
         tmp0 = PCKEV_XORI128_UB(out0, out1);
         tmp1 = PCKEV_XORI128_UB(out2, out3);
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -368,7 +367,7 @@  void ff_put_vp8_epel4_v6_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SRARI_H2_SH(out10, out32, 7);
         SAT_SH2_SH(out10, out32, 7);
         out = PCKEV_XORI128_UB(out10, out32);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         src2110 = src6554;
@@ -416,7 +415,7 @@  void ff_put_vp8_epel8_v6_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SAT_SH4_SH(out0_r, out1_r, out2_r, out3_r, 7);
         tmp0 = PCKEV_XORI128_UB(out0_r, out1_r);
         tmp1 = PCKEV_XORI128_UB(out2_r, out3_r);
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         src10_r = src76_r;
@@ -567,7 +566,7 @@  void ff_put_vp8_epel4_h6v6_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SRARI_H2_SH(tmp0, tmp1, 7);
         SAT_SH2_SH(tmp0, tmp1, 7);
         out = PCKEV_XORI128_UB(tmp0, tmp1);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         hz_out3 = hz_out7;
@@ -651,7 +650,7 @@  void ff_put_vp8_epel8_h6v6_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
         vec0 = PCKEV_XORI128_UB(tmp0, tmp1);
         vec1 = PCKEV_XORI128_UB(tmp2, tmp3);
-        ST8x4_UB(vec0, vec1, dst, dst_stride);
+        ST_D4(vec0, vec1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         hz_out4 = hz_out8;
@@ -702,7 +701,7 @@  static void common_hz_4t_4x4_msa(uint8_t *src, int32_t src_stride,
     SRARI_H2_SH(out0, out1, 7);
     SAT_SH2_SH(out0, out1, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void common_hz_4t_4x8_msa(uint8_t *src, int32_t src_stride,
@@ -735,10 +734,9 @@  static void common_hz_4t_4x8_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_SH(out0, out1, out2, out3, 7);
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
     out = PCKEV_XORI128_UB(out2, out3);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
 }
 
 static void common_hz_4t_4x16_msa(uint8_t *src, int32_t src_stride,
@@ -769,10 +767,10 @@  static void common_hz_4t_4x16_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_SH(out0, out1, out2, out3, 7);
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
     dst += (4 * dst_stride);
     out = PCKEV_XORI128_UB(out2, out3);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
     dst += (4 * dst_stride);
 
     LD_SB8(src, src_stride, src0, src1, src2, src3, src4, src5, src6, src7);
@@ -785,10 +783,10 @@  static void common_hz_4t_4x16_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_SH(out0, out1, out2, out3, 7);
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
     dst += (4 * dst_stride);
     out = PCKEV_XORI128_UB(out2, out3);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 void ff_put_vp8_epel4_h4_msa(uint8_t *dst, ptrdiff_t dst_stride,
@@ -836,7 +834,7 @@  void ff_put_vp8_epel8_h4_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SAT_SH4_SH(out0, out1, out2, out3, 7);
         tmp0 = PCKEV_XORI128_UB(out0, out1);
         tmp1 = PCKEV_XORI128_UB(out2, out3);
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -932,7 +930,7 @@  void ff_put_vp8_epel4_v4_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SRARI_H2_SH(out10, out32, 7);
         SAT_SH2_SH(out10, out32, 7);
         out = PCKEV_XORI128_UB(out10, out32);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -974,7 +972,7 @@  void ff_put_vp8_epel8_v4_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SAT_SH4_SH(out0_r, out1_r, out2_r, out3_r, 7);
         tmp0 = PCKEV_XORI128_UB(out0_r, out1_r);
         tmp1 = PCKEV_XORI128_UB(out2_r, out3_r);
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         src10_r = src98_r;
@@ -1093,7 +1091,7 @@  void ff_put_vp8_epel4_h4v4_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SRARI_H2_SH(tmp0, tmp1, 7);
         SAT_SH2_SH(tmp0, tmp1, 7);
         out = PCKEV_XORI128_UB(tmp0, tmp1);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         hz_out1 = hz_out5;
@@ -1160,7 +1158,7 @@  void ff_put_vp8_epel8_h4v4_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
         out0 = PCKEV_XORI128_UB(tmp0, tmp1);
         out1 = PCKEV_XORI128_UB(tmp2, tmp3);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         vec0 = vec4;
@@ -1240,7 +1238,8 @@  void ff_put_vp8_epel4_h6v4_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SAT_SH2_SH(tmp0, tmp1, 7);
         PCKEV_B2_UB(tmp0, tmp0, tmp1, tmp1, res0, res1);
         XORI_B2_128_UB(res0, res1);
-        ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
+        ST_W2(res0, 0, 1, dst, dst_stride);
+        ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
         dst += (4 * dst_stride);
 
         hz_out1 = hz_out5;
@@ -1316,7 +1315,7 @@  void ff_put_vp8_epel8_h6v4_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
         out0 = PCKEV_XORI128_UB(tmp0, tmp1);
         out1 = PCKEV_XORI128_UB(tmp2, tmp3);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -1391,7 +1390,7 @@  void ff_put_vp8_epel4_h4v6_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SRARI_H2_SH(tmp0, tmp1, 7);
         SAT_SH2_SH(tmp0, tmp1, 7);
         out = PCKEV_XORI128_UB(tmp0, tmp1);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         hz_out3 = hz_out7;
@@ -1464,7 +1463,7 @@  void ff_put_vp8_epel8_h4v6_msa(uint8_t *dst, ptrdiff_t dst_stride,
         SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
         vec0 = PCKEV_XORI128_UB(tmp0, tmp1);
         vec1 = PCKEV_XORI128_UB(tmp2, tmp3);
-        ST8x4_UB(vec0, vec1, dst, dst_stride);
+        ST_D4(vec0, vec1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         hz_out4 = hz_out8;
@@ -1509,7 +1508,8 @@  static void common_hz_2t_4x4_msa(uint8_t *src, int32_t src_stride,
     DOTP_UB2_UH(vec0, vec1, filt0, filt0, vec2, vec3);
     SRARI_H2_UH(vec2, vec3, 7);
     PCKEV_B2_UB(vec2, vec2, vec3, vec3, res0, res1);
-    ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
+    ST_W2(res0, 0, 1, dst, dst_stride);
+    ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
 }
 
 static void common_hz_2t_4x8_msa(uint8_t *src, int32_t src_stride,
@@ -1535,9 +1535,10 @@  static void common_hz_2t_4x8_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_UH(vec4, vec5, vec6, vec7, 7);
     PCKEV_B4_SB(vec4, vec4, vec5, vec5, vec6, vec6, vec7, vec7,
                 res0, res1, res2, res3);
-    ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST4x4_UB(res2, res3, 0, 1, 0, 1, dst, dst_stride);
+    ST_W2(res0, 0, 1, dst, dst_stride);
+    ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
+    ST_W2(res2, 0, 1, dst + 4 * dst_stride, dst_stride);
+    ST_W2(res3, 0, 1, dst + 6 * dst_stride, dst_stride);
 }
 
 void ff_put_vp8_bilinear4_h_msa(uint8_t *dst, ptrdiff_t dst_stride,
@@ -1574,7 +1575,7 @@  static void common_hz_2t_8x4_msa(uint8_t *src, int32_t src_stride,
                 vec0, vec1, vec2, vec3);
     SRARI_H4_UH(vec0, vec1, vec2, vec3, 7);
     PCKEV_B2_SB(vec1, vec0, vec3, vec2, src0, src1);
-    ST8x4_UB(src0, src1, dst, dst_stride);
+    ST_D4(src0, src1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void common_hz_2t_8x8mult_msa(uint8_t *src, int32_t src_stride,
@@ -1604,8 +1605,7 @@  static void common_hz_2t_8x8mult_msa(uint8_t *src, int32_t src_stride,
     src += (4 * src_stride);
 
     PCKEV_B2_SB(vec1, vec0, vec3, vec2, out0, out1);
-    ST8x4_UB(out0, out1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 
     VSHF_B2_UH(src0, src0, src1, src1, mask, mask, vec0, vec1);
     VSHF_B2_UH(src2, src2, src3, src3, mask, mask, vec2, vec3);
@@ -1613,8 +1613,8 @@  static void common_hz_2t_8x8mult_msa(uint8_t *src, int32_t src_stride,
                 vec0, vec1, vec2, vec3);
     SRARI_H4_UH(vec0, vec1, vec2, vec3, 7);
     PCKEV_B2_SB(vec1, vec0, vec3, vec2, out0, out1);
-    ST8x4_UB(out0, out1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
+    dst += (8 * dst_stride);
 
     if (16 == height) {
         LD_SB4(src, src_stride, src0, src1, src2, src3);
@@ -1629,7 +1629,7 @@  static void common_hz_2t_8x8mult_msa(uint8_t *src, int32_t src_stride,
         src += (4 * src_stride);
 
         PCKEV_B2_SB(vec1, vec0, vec3, vec2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 
         VSHF_B2_UH(src0, src0, src1, src1, mask, mask, vec0, vec1);
         VSHF_B2_UH(src2, src2, src3, src3, mask, mask, vec2, vec3);
@@ -1637,7 +1637,7 @@  static void common_hz_2t_8x8mult_msa(uint8_t *src, int32_t src_stride,
                     vec0, vec1, vec2, vec3);
         SRARI_H4_UH(vec0, vec1, vec2, vec3, 7);
         PCKEV_B2_SB(vec1, vec0, vec3, vec2, out0, out1);
-        ST8x4_UB(out0, out1, dst + 4 * dst_stride, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
     }
 }
 
@@ -1745,7 +1745,7 @@  static void common_vt_2t_4x4_msa(uint8_t *src, int32_t src_stride,
     SRARI_H2_UH(tmp0, tmp1, 7);
     SAT_UH2_UH(tmp0, tmp1, 7);
     src2110 = __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-    ST4x4_UB(src2110, src2110, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(src2110, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void common_vt_2t_4x8_msa(uint8_t *src, int32_t src_stride,
@@ -1779,8 +1779,7 @@  static void common_vt_2t_4x8_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, src2110, src4332);
-    ST4x4_UB(src2110, src2110, 0, 1, 2, 3, dst, dst_stride);
-    ST4x4_UB(src4332, src4332, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
+    ST_W8(src2110, src4332, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
 }
 
 void ff_put_vp8_bilinear4_v_msa(uint8_t *dst, ptrdiff_t dst_stride,
@@ -1817,7 +1816,7 @@  static void common_vt_2t_8x4_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-    ST8x4_UB(out0, out1, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void common_vt_2t_8x8mult_msa(uint8_t *src, int32_t src_stride,
@@ -1851,16 +1850,15 @@  static void common_vt_2t_8x8mult_msa(uint8_t *src, int32_t src_stride,
         SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
         SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
         PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
-        dst += (4 * dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 
         DOTP_UB4_UH(vec4, vec5, vec6, vec7, filt0, filt0, filt0, filt0,
                     tmp0, tmp1, tmp2, tmp3);
         SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
         SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
         PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
-        dst += (4 * dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
+        dst += (8 * dst_stride);
 
         src0 = src8;
     }
@@ -1964,7 +1962,8 @@  static void common_hv_2ht_2vt_4x4_msa(uint8_t *src, int32_t src_stride,
     SRARI_H2_UH(tmp0, tmp1, 7);
     SAT_UH2_UH(tmp0, tmp1, 7);
     PCKEV_B2_UB(tmp0, tmp0, tmp1, tmp1, res0, res1);
-    ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
+    ST_W2(res0, 0, 1, dst, dst_stride);
+    ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
 }
 
 static void common_hv_2ht_2vt_4x8_msa(uint8_t *src, int32_t src_stride,
@@ -2008,9 +2007,10 @@  static void common_hv_2ht_2vt_4x8_msa(uint8_t *src, int32_t src_stride,
     SAT_UH4_UH(vec4, vec5, vec6, vec7, 7);
     PCKEV_B4_SB(vec4, vec4, vec5, vec5, vec6, vec6, vec7, vec7,
                 res0, res1, res2, res3);
-    ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST4x4_UB(res2, res3, 0, 1, 0, 1, dst, dst_stride);
+    ST_W2(res0, 0, 1, dst, dst_stride);
+    ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
+    ST_W2(res2, 0, 1, dst + 4 * dst_stride, dst_stride);
+    ST_W2(res3, 0, 1, dst + 6 * dst_stride, dst_stride);
 }
 
 void ff_put_vp8_bilinear4_hv_msa(uint8_t *dst, ptrdiff_t dst_stride,
@@ -2070,7 +2070,7 @@  static void common_hv_2ht_2vt_8x4_msa(uint8_t *src, int32_t src_stride,
     SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-    ST8x4_UB(out0, out1, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void common_hv_2ht_2vt_8x8mult_msa(uint8_t *src, int32_t src_stride,
@@ -2127,8 +2127,7 @@  static void common_hv_2ht_2vt_8x8mult_msa(uint8_t *src, int32_t src_stride,
         SRARI_H2_UH(tmp3, tmp4, 7);
         SAT_UH2_UH(tmp3, tmp4, 7);
         PCKEV_B2_SB(tmp2, tmp1, tmp4, tmp3, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
-        dst += (4 * dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 
         hz_out1 = HORIZ_2TAP_FILT_UH(src1, src1, mask, filt_hz, 7);
         vec0 = (v16u8) __msa_ilvev_b((v16i8) hz_out1, (v16i8) hz_out0);
@@ -2149,8 +2148,8 @@  static void common_hv_2ht_2vt_8x8mult_msa(uint8_t *src, int32_t src_stride,
         SRARI_H4_UH(tmp5, tmp6, tmp7, tmp8, 7);
         SAT_UH4_UH(tmp5, tmp6, tmp7, tmp8, 7);
         PCKEV_B2_SB(tmp6, tmp5, tmp8, tmp7, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
-        dst += (4 * dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
+        dst += (8 * dst_stride);
     }
 }
 
diff --git a/libavcodec/mips/vp9_idct_msa.c b/libavcodec/mips/vp9_idct_msa.c
index bd762f2..1f32770 100644
--- a/libavcodec/mips/vp9_idct_msa.c
+++ b/libavcodec/mips/vp9_idct_msa.c
@@ -241,7 +241,7 @@  static const int32_t sinpi_4_9 = 15212;
          res0_m, res1_m, res2_m, res3_m);                         \
     CLIP_SH4_0_255(res0_m, res1_m, res2_m, res3_m);               \
     PCKEV_B2_SB(res1_m, res0_m, res3_m, res2_m, tmp0_m, tmp1_m);  \
-    ST8x4_UB(tmp0_m, tmp1_m, dst_m, dst_stride);                  \
+    ST_D4(tmp0_m, tmp1_m, 0, 1, 0, 1, dst_m, dst_stride);         \
 }
 
 #define VP9_IDCT4x4(in0, in1, in2, in3, out0, out1, out2, out3)       \
@@ -364,7 +364,10 @@  static void vp9_idct4x4_colcol_addblk_msa(int16_t *input, uint8_t *dst,
     v8i16 zero = { 0 };
 
     /* load vector elements of 4x4 block */
-    LD4x4_SH(input, in0, in1, in2, in3);
+    in0 = LD_SH(input);
+    in2 = LD_SH(input + 8);
+    in1 = (v8i16) __msa_ilvl_d((v2i64) in0, (v2i64) in0);
+    in3 = (v8i16) __msa_ilvl_d((v2i64) in2, (v2i64) in2);
     ST_SH2(zero, zero, input, 8);
     /* rows */
     VP9_IDCT4x4(in0, in1, in2, in3, in0, in1, in2, in3);
@@ -383,7 +386,10 @@  static void vp9_iadst4x4_colcol_addblk_msa(int16_t *input, uint8_t *dst,
     v8i16 zero = { 0 };
 
     /* load vector elements of 4x4 block */
-    LD4x4_SH(input, in0, in1, in2, in3);
+    in0 = LD_SH(input);
+    in2 = LD_SH(input + 8);
+    in1 = (v8i16) __msa_ilvl_d((v2i64) in0, (v2i64) in0);
+    in3 = (v8i16) __msa_ilvl_d((v2i64) in2, (v2i64) in2);
     ST_SH2(zero, zero, input, 8);
     /* rows */
     VP9_IADST4x4(in0, in1, in2, in3, in0, in1, in2, in3);
@@ -402,7 +408,10 @@  static void vp9_iadst_idct_4x4_add_msa(int16_t *input, uint8_t *dst,
     v8i16 zero = { 0 };
 
     /* load vector elements of 4x4 block */
-    LD4x4_SH(input, in0, in1, in2, in3);
+    in0 = LD_SH(input);
+    in2 = LD_SH(input + 8);
+    in1 = (v8i16) __msa_ilvl_d((v2i64) in0, (v2i64) in0);
+    in3 = (v8i16) __msa_ilvl_d((v2i64) in2, (v2i64) in2);
     ST_SH2(zero, zero, input, 8);
     /* cols */
     VP9_IADST4x4(in0, in1, in2, in3, in0, in1, in2, in3);
@@ -421,7 +430,10 @@  static void vp9_idct_iadst_4x4_add_msa(int16_t *input, uint8_t *dst,
     v8i16 zero = { 0 };
 
     /* load vector elements of 4x4 block */
-    LD4x4_SH(input, in0, in1, in2, in3);
+    in0 = LD_SH(input);
+    in2 = LD_SH(input + 8);
+    in1 = (v8i16) __msa_ilvl_d((v2i64) in0, (v2i64) in0);
+    in3 = (v8i16) __msa_ilvl_d((v2i64) in2, (v2i64) in2);
     ST_SH2(zero, zero, input, 8);
     /* cols */
     VP9_IDCT4x4(in0, in1, in2, in3, in0, in1, in2, in3);
@@ -753,13 +765,13 @@  static void vp9_iadst8x8_colcol_addblk_msa(int16_t *input, uint8_t *dst,
     res0 += out0;
     res0 = CLIP_SH_0_255(res0);
     res0 = (v8i16) __msa_pckev_b((v16i8) res0, (v16i8) res0);
-    ST8x1_UB(res0, dst);
+    ST_D1(res0, 0, dst);
 
     res7 = (v8i16) __msa_ilvr_b((v16i8) zero, (v16i8) dst7);
     res7 += out7;
     res7 = CLIP_SH_0_255(res7);
     res7 = (v8i16) __msa_pckev_b((v16i8) res7, (v16i8) res7);
-    ST8x1_UB(res7, dst + 7 * dst_stride);
+    ST_D1(res7, 0, dst + 7 * dst_stride);
 
     cnst1 = __msa_fill_h(cospi_24_64);
     cnst0 = __msa_fill_h(cospi_8_64);
@@ -782,8 +794,8 @@  static void vp9_iadst8x8_colcol_addblk_msa(int16_t *input, uint8_t *dst,
     ADD2(res1, out1, res6, out6, res1, res6);
     CLIP_SH2_0_255(res1, res6);
     PCKEV_B2_SH(res1, res1, res6, res6, res1, res6);
-    ST8x1_UB(res1, dst + dst_stride);
-    ST8x1_UB(res6, dst + 6 * dst_stride);
+    ST_D1(res1, 0, dst + dst_stride);
+    ST_D1(res6, 0, dst + 6 * dst_stride);
 
     cnst0 = __msa_fill_h(cospi_16_64);
     cnst1 = -cnst0;
@@ -801,8 +813,8 @@  static void vp9_iadst8x8_colcol_addblk_msa(int16_t *input, uint8_t *dst,
     ADD2(res3, out3, res4, out4, res3, res4);
     CLIP_SH2_0_255(res3, res4);
     PCKEV_B2_SH(res3, res3, res4, res4, res3, res4);
-    ST8x1_UB(res3, dst + 3 * dst_stride);
-    ST8x1_UB(res4, dst + 4 * dst_stride);
+    ST_D1(res3, 0, dst + 3 * dst_stride);
+    ST_D1(res4, 0, dst + 4 * dst_stride);
 
     out2 = VP9_DOT_SHIFT_RIGHT_PCK_H(temp2, temp3, cnst0);
     out5 = VP9_DOT_SHIFT_RIGHT_PCK_H(temp2, temp3, cnst1);
@@ -814,8 +826,8 @@  static void vp9_iadst8x8_colcol_addblk_msa(int16_t *input, uint8_t *dst,
     ADD2(res2, out2, res5, out5, res2, res5);
     CLIP_SH2_0_255(res2, res5);
     PCKEV_B2_SH(res2, res2, res5, res5, res2, res5);
-    ST8x1_UB(res2, dst + 2 * dst_stride);
-    ST8x1_UB(res5, dst + 5 * dst_stride);
+    ST_D1(res2, 0, dst + 2 * dst_stride);
+    ST_D1(res5, 0, dst + 5 * dst_stride);
 }
 
 static void vp9_iadst_idct_8x8_add_msa(int16_t *input, uint8_t *dst,
@@ -1354,8 +1366,8 @@  static void vp9_iadst16_1d_columns_addblk_msa(int16_t *input, uint8_t *dst,
     ADD2(res0, out0, res1, out1, res0, res1);
     CLIP_SH2_0_255(res0, res1);
     PCKEV_B2_SH(res0, res0, res1, res1, res0, res1);
-    ST8x1_UB(res0, dst);
-    ST8x1_UB(res1, dst + 15 * dst_stride);
+    ST_D1(res0, 0, dst);
+    ST_D1(res1, 0, dst + 15 * dst_stride);
 
     k0 = VP9_SET_COSPI_PAIR(cospi_12_64, cospi_20_64);
     k1 = VP9_SET_COSPI_PAIR(-cospi_20_64, cospi_12_64);
@@ -1371,8 +1383,8 @@  static void vp9_iadst16_1d_columns_addblk_msa(int16_t *input, uint8_t *dst,
     ADD2(res8, out8, res9, out9, res8, res9);
     CLIP_SH2_0_255(res8, res9);
     PCKEV_B2_SH(res8, res8, res9, res9, res8, res9);
-    ST8x1_UB(res8, dst + dst_stride);
-    ST8x1_UB(res9, dst + 14 * dst_stride);
+    ST_D1(res8, 0, dst + dst_stride);
+    ST_D1(res9, 0, dst + 14 * dst_stride);
 
     k0 = VP9_SET_COSPI_PAIR(cospi_8_64, cospi_24_64);
     k1 = VP9_SET_COSPI_PAIR(cospi_24_64, -cospi_8_64);
@@ -1386,8 +1398,8 @@  static void vp9_iadst16_1d_columns_addblk_msa(int16_t *input, uint8_t *dst,
     ADD2(res4, out4, res5, out5, res4, res5);
     CLIP_SH2_0_255(res4, res5);
     PCKEV_B2_SH(res4, res4, res5, res5, res4, res5);
-    ST8x1_UB(res4, dst + 3 * dst_stride);
-    ST8x1_UB(res5, dst + 12 * dst_stride);
+    ST_D1(res4, 0, dst + 3 * dst_stride);
+    ST_D1(res5, 0, dst + 12 * dst_stride);
 
     VP9_MADD_BF(h1, h3, h5, h7, k0, k1, k2, k0, out12, out14, out13, out15);
     out13 = -out13;
@@ -1398,8 +1410,8 @@  static void vp9_iadst16_1d_columns_addblk_msa(int16_t *input, uint8_t *dst,
     ADD2(res12, out12, res13, out13, res12, res13);
     CLIP_SH2_0_255(res12, res13);
     PCKEV_B2_SH(res12, res12, res13, res13, res12, res13);
-    ST8x1_UB(res12, dst + 2 * dst_stride);
-    ST8x1_UB(res13, dst + 13 * dst_stride);
+    ST_D1(res12, 0, dst + 2 * dst_stride);
+    ST_D1(res13, 0, dst + 13 * dst_stride);
 
     k0 = VP9_SET_COSPI_PAIR(cospi_16_64, cospi_16_64);
     k3 = VP9_SET_COSPI_PAIR(-cospi_16_64, cospi_16_64);
@@ -1411,8 +1423,8 @@  static void vp9_iadst16_1d_columns_addblk_msa(int16_t *input, uint8_t *dst,
     ADD2(res6, out6, res7, out7, res6, res7);
     CLIP_SH2_0_255(res6, res7);
     PCKEV_B2_SH(res6, res6, res7, res7, res6, res7);
-    ST8x1_UB(res6, dst + 4 * dst_stride);
-    ST8x1_UB(res7, dst + 11 * dst_stride);
+    ST_D1(res6, 0, dst + 4 * dst_stride);
+    ST_D1(res7, 0, dst + 11 * dst_stride);
 
     VP9_MADD_SHORT(out10, out11, k0, k3, out10, out11);
     SRARI_H2_SH(out10, out11, 6);
@@ -1422,8 +1434,8 @@  static void vp9_iadst16_1d_columns_addblk_msa(int16_t *input, uint8_t *dst,
     ADD2(res10, out10, res11, out11, res10, res11);
     CLIP_SH2_0_255(res10, res11);
     PCKEV_B2_SH(res10, res10, res11, res11, res10, res11);
-    ST8x1_UB(res10, dst + 6 * dst_stride);
-    ST8x1_UB(res11, dst + 9 * dst_stride);
+    ST_D1(res10, 0, dst + 6 * dst_stride);
+    ST_D1(res11, 0, dst + 9 * dst_stride);
 
     k1 = VP9_SET_COSPI_PAIR(-cospi_16_64, -cospi_16_64);
     k2 = VP9_SET_COSPI_PAIR(cospi_16_64, -cospi_16_64);
@@ -1435,8 +1447,8 @@  static void vp9_iadst16_1d_columns_addblk_msa(int16_t *input, uint8_t *dst,
     ADD2(res2, out2, res3, out3, res2, res3);
     CLIP_SH2_0_255(res2, res3);
     PCKEV_B2_SH(res2, res2, res3, res3, res2, res3);
-    ST8x1_UB(res2, dst + 7 * dst_stride);
-    ST8x1_UB(res3, dst + 8 * dst_stride);
+    ST_D1(res2, 0, dst + 7 * dst_stride);
+    ST_D1(res3, 0, dst + 8 * dst_stride);
 
     VP9_MADD_SHORT(out14, out15, k1, k2, out14, out15);
     SRARI_H2_SH(out14, out15, 6);
@@ -1446,8 +1458,8 @@  static void vp9_iadst16_1d_columns_addblk_msa(int16_t *input, uint8_t *dst,
     ADD2(res14, out14, res15, out15, res14, res15);
     CLIP_SH2_0_255(res14, res15);
     PCKEV_B2_SH(res14, res14, res15, res15, res14, res15);
-    ST8x1_UB(res14, dst + 5 * dst_stride);
-    ST8x1_UB(res15, dst + 10 * dst_stride);
+    ST_D1(res14, 0, dst + 5 * dst_stride);
+    ST_D1(res15, 0, dst + 10 * dst_stride);
 }
 
 static void vp9_iadst16x16_colcol_addblk_msa(int16_t *input, uint8_t *dst,
diff --git a/libavcodec/mips/vp9_intra_msa.c b/libavcodec/mips/vp9_intra_msa.c
index 54cf0ae..97cf212 100644
--- a/libavcodec/mips/vp9_intra_msa.c
+++ b/libavcodec/mips/vp9_intra_msa.c
@@ -378,7 +378,8 @@  void ff_tm_4x4_msa(uint8_t *dst, ptrdiff_t dst_stride,
     IPRED_SUBS_UH2_UH(src_top_left, src_top_left, vec2, vec3);
     SAT_UH4_UH(vec0, vec1, vec2, vec3, 7);
     PCKEV_B2_SB(vec1, vec0, vec3, vec2, tmp0, tmp1);
-    ST4x4_UB(tmp0, tmp1, 0, 2, 0, 2, dst, dst_stride);
+    ST_W2(tmp0, 0, 2, dst, dst_stride);
+    ST_W2(tmp1, 0, 2, dst + 2 * dst_stride, dst_stride);
 }
 
 void ff_tm_8x8_msa(uint8_t *dst, ptrdiff_t dst_stride,
@@ -409,7 +410,7 @@  void ff_tm_8x8_msa(uint8_t *dst, ptrdiff_t dst_stride,
         IPRED_SUBS_UH2_UH(src_top_left, src_top_left, vec2, vec3);
         SAT_UH4_UH(vec0, vec1, vec2, vec3, 7);
         PCKEV_B2_SB(vec1, vec0, vec3, vec2, tmp0, tmp1);
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
diff --git a/libavcodec/mips/vp9_lpf_msa.c b/libavcodec/mips/vp9_lpf_msa.c
index c82a9e9..2450c74 100644
--- a/libavcodec/mips/vp9_lpf_msa.c
+++ b/libavcodec/mips/vp9_lpf_msa.c
@@ -1219,9 +1219,7 @@  void ff_loop_filter_h_4_8_msa(uint8_t *src, ptrdiff_t pitch,
     ILVRL_H2_SH(vec1, vec0, vec2, vec3);
 
     src -= 2;
-    ST4x4_UB(vec2, vec2, 0, 1, 2, 3, src, pitch);
-    src += 4 * pitch;
-    ST4x4_UB(vec3, vec3, 0, 1, 2, 3, src, pitch);
+    ST_W8(vec2, vec3, 0, 1, 2, 3, 0, 1, 2, 3, src, pitch);
 }
 
 void ff_loop_filter_h_44_16_msa(uint8_t *src, ptrdiff_t pitch,
@@ -1266,9 +1264,8 @@  void ff_loop_filter_h_44_16_msa(uint8_t *src, ptrdiff_t pitch,
 
     src -= 2;
 
-    ST4x8_UB(tmp2, tmp3, src, pitch);
-    src += (8 * pitch);
-    ST4x8_UB(tmp4, tmp5, src, pitch);
+    ST_W8(tmp2, tmp3, 0, 1, 2, 3, 0, 1, 2, 3, src, pitch);
+    ST_W8(tmp4, tmp5, 0, 1, 2, 3, 0, 1, 2, 3, src + 8 * pitch, pitch);
 }
 
 void ff_loop_filter_h_8_8_msa(uint8_t *src, ptrdiff_t pitch,
@@ -1313,9 +1310,7 @@  void ff_loop_filter_h_8_8_msa(uint8_t *src, ptrdiff_t pitch,
         ILVRL_H2_SH(vec1, vec0, vec2, vec3);
 
         src -= 2;
-        ST4x4_UB(vec2, vec2, 0, 1, 2, 3, src, pitch);
-        src += 4 * pitch;
-        ST4x4_UB(vec3, vec3, 0, 1, 2, 3, src, pitch);
+        ST_W8(vec2, vec3, 0, 1, 2, 3, 0, 1, 2, 3, src, pitch);
     } else {
         ILVR_B8_UH(zero, p3, zero, p2, zero, p1, zero, p0, zero, q0, zero, q1,
                    zero, q2, zero, q3, p3_r, p2_r, p1_r, p0_r, q0_r, q1_r, q2_r,
@@ -1343,11 +1338,11 @@  void ff_loop_filter_h_8_8_msa(uint8_t *src, ptrdiff_t pitch,
         vec4 = (v8i16) __msa_ilvr_b((v16i8) q2, (v16i8) q1);
 
         src -= 3;
-        ST4x4_UB(vec2, vec2, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec4, 0, src + 4, pitch);
+        ST_W4(vec2, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec4, 0, 1, 2, 3, src + 4, pitch);
         src += (4 * pitch);
-        ST4x4_UB(vec3, vec3, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec4, 4, src + 4, pitch);
+        ST_W4(vec3, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec4, 4, 5, 6, 7, src + 4, pitch);
     }
 }
 
@@ -1410,9 +1405,8 @@  void ff_loop_filter_h_88_16_msa(uint8_t *src, ptrdiff_t pitch,
         ILVRL_H2_SH(vec1, vec0, vec4, vec5);
 
         src -= 2;
-        ST4x8_UB(vec2, vec3, src, pitch);
-        src += 8 * pitch;
-        ST4x8_UB(vec4, vec5, src, pitch);
+        ST_W8(vec2, vec3, 0, 1, 2, 3, 0, 1, 2, 3, src, pitch);
+        ST_W8(vec4, vec5, 0, 1, 2, 3, 0, 1, 2, 3, src + 8 * pitch, pitch);
     } else {
         ILVR_B8_UH(zero, p3, zero, p2, zero, p1, zero, p0, zero, q0, zero, q1,
                    zero, q2, zero, q3, p3_r, p2_r, p1_r, p0_r, q0_r, q1_r, q2_r,
@@ -1451,17 +1445,17 @@  void ff_loop_filter_h_88_16_msa(uint8_t *src, ptrdiff_t pitch,
         ILVRL_B2_SH(q2, q1, vec2, vec5);
 
         src -= 3;
-        ST4x4_UB(vec3, vec3, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec2, 0, src + 4, pitch);
+        ST_W4(vec3, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec2, 0, 1, 2, 3, src + 4, pitch);
         src += (4 * pitch);
-        ST4x4_UB(vec4, vec4, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec2, 4, src + 4, pitch);
+        ST_W4(vec4, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec2, 4, 5, 6, 7, src + 4, pitch);
         src += (4 * pitch);
-        ST4x4_UB(vec6, vec6, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec5, 0, src + 4, pitch);
+        ST_W4(vec6, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec5, 0, 1, 2, 3, src + 4, pitch);
         src += (4 * pitch);
-        ST4x4_UB(vec7, vec7, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec5, 4, src + 4, pitch);
+        ST_W4(vec7, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec5, 4, 5, 6, 7, src + 4, pitch);
     }
 }
 
@@ -1523,9 +1517,8 @@  void ff_loop_filter_h_84_16_msa(uint8_t *src, ptrdiff_t pitch,
         ILVRL_H2_SH(vec1, vec0, vec4, vec5);
 
         src -= 2;
-        ST4x8_UB(vec2, vec3, src, pitch);
-        src += 8 * pitch;
-        ST4x8_UB(vec4, vec5, src, pitch);
+        ST_W8(vec2, vec3, 0, 1, 2, 3, 0, 1, 2, 3, src, pitch);
+        ST_W8(vec4, vec5, 0, 1, 2, 3, 0, 1, 2, 3, src + 8 * pitch, pitch);
     } else {
         ILVR_B8_UH(zero, p3, zero, p2, zero, p1, zero, p0, zero, q0, zero, q1,
                    zero, q2, zero, q3, p3_r, p2_r, p1_r, p0_r, q0_r, q1_r, q2_r,
@@ -1555,17 +1548,17 @@  void ff_loop_filter_h_84_16_msa(uint8_t *src, ptrdiff_t pitch,
         ILVRL_B2_SH(q2, q1, vec2, vec5);
 
         src -= 3;
-        ST4x4_UB(vec3, vec3, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec2, 0, src + 4, pitch);
+        ST_W4(vec3, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec2, 0, 1, 2, 3, src + 4, pitch);
         src += (4 * pitch);
-        ST4x4_UB(vec4, vec4, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec2, 4, src + 4, pitch);
+        ST_W4(vec4, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec2, 4, 5, 6, 7, src + 4, pitch);
         src += (4 * pitch);
-        ST4x4_UB(vec6, vec6, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec5, 0, src + 4, pitch);
+        ST_W4(vec6, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec5, 0, 1, 2, 3, src + 4, pitch);
         src += (4 * pitch);
-        ST4x4_UB(vec7, vec7, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec5, 4, src + 4, pitch);
+        ST_W4(vec7, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec5, 4, 5, 6, 7, src + 4, pitch);
     }
 }
 
@@ -1627,9 +1620,8 @@  void ff_loop_filter_h_48_16_msa(uint8_t *src, ptrdiff_t pitch,
         ILVRL_H2_SH(vec1, vec0, vec4, vec5);
 
         src -= 2;
-        ST4x8_UB(vec2, vec3, src, pitch);
-        src += 8 * pitch;
-        ST4x8_UB(vec4, vec5, src, pitch);
+        ST_W8(vec2, vec3, 0, 1, 2, 3, 0, 1, 2, 3, src, pitch);
+        ST_W8(vec4, vec5, 0, 1, 2, 3, 0, 1, 2, 3, src + 8 * pitch, pitch);
     } else {
         ILVL_B4_UH(zero, p3, zero, p2, zero, p1, zero, p0, p3_l, p2_l, p1_l,
                    p0_l);
@@ -1661,17 +1653,17 @@  void ff_loop_filter_h_48_16_msa(uint8_t *src, ptrdiff_t pitch,
         ILVRL_B2_SH(q2, q1, vec2, vec5);
 
         src -= 3;
-        ST4x4_UB(vec3, vec3, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec2, 0, src + 4, pitch);
+        ST_W4(vec3, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec2, 0, 1, 2, 3, src + 4, pitch);
         src += (4 * pitch);
-        ST4x4_UB(vec4, vec4, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec2, 4, src + 4, pitch);
+        ST_W4(vec4, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec2, 4, 5, 6, 7, src + 4, pitch);
         src += (4 * pitch);
-        ST4x4_UB(vec6, vec6, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec5, 0, src + 4, pitch);
+        ST_W4(vec6, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec5, 0, 1, 2, 3, src + 4, pitch);
         src += (4 * pitch);
-        ST4x4_UB(vec7, vec7, 0, 1, 2, 3, src, pitch);
-        ST2x4_UB(vec5, 4, src + 4, pitch);
+        ST_W4(vec7, 0, 1, 2, 3, src, pitch);
+        ST_H4(vec5, 4, 5, 6, 7, src + 4, pitch);
     }
 }
 
@@ -1811,7 +1803,7 @@  static int32_t vp9_vt_lpf_t4_and_t8_8w(uint8_t *src, uint8_t *filter48,
     if (__msa_test_bz_v(flat)) {
         ILVR_B2_SH(p0_out, p1_out, q1_out, q0_out, vec0, vec1);
         ILVRL_H2_SH(vec1, vec0, vec2, vec3);
-        ST4x8_UB(vec2, vec3, (src_org - 2), pitch_org);
+        ST_W8(vec2, vec3, 0, 1, 2, 3, 0, 1, 2, 3, (src_org - 2), pitch_org);
         return 1;
     } else {
         ILVR_B8_UH(zero, p3, zero, p2, zero, p1, zero, p0, zero, q0, zero, q1,
@@ -1878,11 +1870,11 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         vec2 = (v8i16) __msa_ilvr_b((v16i8) q2, (v16i8) q1);
 
         src_org -= 3;
-        ST4x4_UB(vec3, vec3, 0, 1, 2, 3, src_org, pitch);
-        ST2x4_UB(vec2, 0, (src_org + 4), pitch);
+        ST_W4(vec3, 0, 1, 2, 3, src_org, pitch);
+        ST_H4(vec2, 0, 1, 2, 3, (src_org + 4), pitch);
         src_org += (4 * pitch);
-        ST4x4_UB(vec4, vec4, 0, 1, 2, 3, src_org, pitch);
-        ST2x4_UB(vec2, 4, (src_org + 4), pitch);
+        ST_W4(vec4, 0, 1, 2, 3, src_org, pitch);
+        ST_H4(vec2, 4, 5, 6, 7, (src_org + 4), pitch);
 
         return 1;
     } else {
@@ -1908,7 +1900,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         p6 = __msa_bmnz_v(p6, (v16u8) r_out, flat2);
-        ST8x1_UB(p6, src);
+        ST_D1(p6, 0, src);
         src += 16;
 
         /* p5 */
@@ -1920,7 +1912,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         p5 = __msa_bmnz_v(p5, (v16u8) r_out, flat2);
-        ST8x1_UB(p5, src);
+        ST_D1(p5, 0, src);
         src += 16;
 
         /* p4 */
@@ -1932,7 +1924,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         p4 = __msa_bmnz_v(p4, (v16u8) r_out, flat2);
-        ST8x1_UB(p4, src);
+        ST_D1(p4, 0, src);
         src += 16;
 
         /* p3 */
@@ -1944,7 +1936,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         p3 = __msa_bmnz_v(p3, (v16u8) r_out, flat2);
-        ST8x1_UB(p3, src);
+        ST_D1(p3, 0, src);
         src += 16;
 
         /* p2 */
@@ -1957,7 +1949,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         filter8 = __msa_bmnz_v(filter8, (v16u8) r_out, flat2);
-        ST8x1_UB(filter8, src);
+        ST_D1(filter8, 0, src);
         src += 16;
 
         /* p1 */
@@ -1970,7 +1962,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         filter8 = __msa_bmnz_v(filter8, (v16u8) r_out, flat2);
-        ST8x1_UB(filter8, src);
+        ST_D1(filter8, 0, src);
         src += 16;
 
         /* p0 */
@@ -1983,7 +1975,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         filter8 = __msa_bmnz_v(filter8, (v16u8) r_out, flat2);
-        ST8x1_UB(filter8, src);
+        ST_D1(filter8, 0, src);
         src += 16;
 
         /* q0 */
@@ -1996,7 +1988,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         filter8 = __msa_bmnz_v(filter8, (v16u8) r_out, flat2);
-        ST8x1_UB(filter8, src);
+        ST_D1(filter8, 0, src);
         src += 16;
 
         /* q1 */
@@ -2008,7 +2000,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         filter8 = __msa_bmnz_v(filter8, (v16u8) r_out, flat2);
-        ST8x1_UB(filter8, src);
+        ST_D1(filter8, 0, src);
         src += 16;
 
         /* q2 */
@@ -2020,7 +2012,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         filter8 = __msa_bmnz_v(filter8, (v16u8) r_out, flat2);
-        ST8x1_UB(filter8, src);
+        ST_D1(filter8, 0, src);
         src += 16;
 
         /* q3 */
@@ -2031,7 +2023,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         q3 = __msa_bmnz_v(q3, (v16u8) r_out, flat2);
-        ST8x1_UB(q3, src);
+        ST_D1(q3, 0, src);
         src += 16;
 
         /* q4 */
@@ -2042,7 +2034,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         q4 = __msa_bmnz_v(q4, (v16u8) r_out, flat2);
-        ST8x1_UB(q4, src);
+        ST_D1(q4, 0, src);
         src += 16;
 
         /* q5 */
@@ -2053,7 +2045,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         q5 = __msa_bmnz_v(q5, (v16u8) r_out, flat2);
-        ST8x1_UB(q5, src);
+        ST_D1(q5, 0, src);
         src += 16;
 
         /* q6 */
@@ -2064,7 +2056,7 @@  static int32_t vp9_vt_lpf_t16_8w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitch
         r_out = __msa_srari_h((v8i16) tmp1_r, 4);
         r_out = (v8i16) __msa_pckev_b((v16i8) r_out, (v16i8) r_out);
         q6 = __msa_bmnz_v(q6, (v16u8) r_out, flat2);
-        ST8x1_UB(q6, src);
+        ST_D1(q6, 0, src);
 
         return 0;
     }
@@ -2137,9 +2129,8 @@  static int32_t vp9_vt_lpf_t4_and_t8_16w(uint8_t *src, uint8_t *filter48,
         ILVRL_H2_SH(vec1, vec0, vec4, vec5);
 
         src_org -= 2;
-        ST4x8_UB(vec2, vec3, src_org, pitch);
-        src_org += 8 * pitch;
-        ST4x8_UB(vec4, vec5, src_org, pitch);
+        ST_W8(vec2, vec3, 0, 1, 2, 3, 0, 1, 2, 3, src_org, pitch);
+        ST_W8(vec4, vec5, 0, 1, 2, 3, 0, 1, 2, 3, src_org + 8 * pitch, pitch);
 
         return 1;
     } else {
@@ -2218,17 +2209,17 @@  static int32_t vp9_vt_lpf_t16_16w(uint8_t *src, uint8_t *src_org, ptrdiff_t pitc
         ILVRL_B2_SH(q2, q1, vec2, vec5);
 
         src_org -= 3;
-        ST4x4_UB(vec3, vec3, 0, 1, 2, 3, src_org, pitch);
-        ST2x4_UB(vec2, 0, (src_org + 4), pitch);
+        ST_W4(vec3, 0, 1, 2, 3, src_org, pitch);
+        ST_H4(vec2, 0, 1, 2, 3, (src_org + 4), pitch);
         src_org += (4 * pitch);
-        ST4x4_UB(vec4, vec4, 0, 1, 2, 3, src_org, pitch);
-        ST2x4_UB(vec2, 4, (src_org + 4), pitch);
+        ST_W4(vec4, 0, 1, 2, 3, src_org, pitch);
+        ST_H4(vec2, 4, 5, 6, 7, (src_org + 4), pitch);
         src_org += (4 * pitch);
-        ST4x4_UB(vec6, vec6, 0, 1, 2, 3, src_org, pitch);
-        ST2x4_UB(vec5, 0, (src_org + 4), pitch);
+        ST_W4(vec6, 0, 1, 2, 3, src_org, pitch);
+        ST_H4(vec5, 0, 1, 2, 3, (src_org + 4), pitch);
         src_org += (4 * pitch);
-        ST4x4_UB(vec7, vec7, 0, 1, 2, 3, src_org, pitch);
-        ST2x4_UB(vec5, 4, (src_org + 4), pitch);
+        ST_W4(vec7, 0, 1, 2, 3, src_org, pitch);
+        ST_H4(vec5, 4, 5, 6, 7, (src_org + 4), pitch);
 
         return 1;
     } else {
diff --git a/libavcodec/mips/vp9_mc_msa.c b/libavcodec/mips/vp9_mc_msa.c
index 749e8cb..1d8a892 100644
--- a/libavcodec/mips/vp9_mc_msa.c
+++ b/libavcodec/mips/vp9_mc_msa.c
@@ -153,7 +153,7 @@  static const int8_t vp9_bilinear_filters_msa[15][2] = {
                                                               \
     PCKEV_B2_UB(in1, in0, in3, in2, tmp0_m, tmp1_m);          \
     AVER_UB2_UB(tmp0_m, dst0, tmp1_m, dst1, tmp0_m, tmp1_m);  \
-    ST8x4_UB(tmp0_m, tmp1_m, pdst_m, stride);                 \
+    ST_D4(tmp0_m, tmp1_m, 0, 1, 0, 1, pdst_m, stride);        \
 }
 
 static void common_hz_8t_4x4_msa(const uint8_t *src, int32_t src_stride,
@@ -182,7 +182,7 @@  static void common_hz_8t_4x4_msa(const uint8_t *src, int32_t src_stride,
     SRARI_H2_SH(out0, out1, 7);
     SAT_SH2_SH(out0, out1, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void common_hz_8t_4x8_msa(const uint8_t *src, int32_t src_stride,
@@ -217,10 +217,9 @@  static void common_hz_8t_4x8_msa(const uint8_t *src, int32_t src_stride,
     SRARI_H4_SH(out0, out1, out2, out3, 7);
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     out = PCKEV_XORI128_UB(out0, out1);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
     out = PCKEV_XORI128_UB(out2, out3);
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
 }
 
 static void common_hz_8t_4w_msa(const uint8_t *src, int32_t src_stride,
@@ -262,7 +261,7 @@  static void common_hz_8t_8x4_msa(const uint8_t *src, int32_t src_stride,
     SAT_SH4_SH(out0, out1, out2, out3, 7);
     tmp0 = PCKEV_XORI128_UB(out0, out1);
     tmp1 = PCKEV_XORI128_UB(out2, out3);
-    ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+    ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void common_hz_8t_8x8mult_msa(const uint8_t *src, int32_t src_stride,
@@ -296,7 +295,7 @@  static void common_hz_8t_8x8mult_msa(const uint8_t *src, int32_t src_stride,
         SAT_SH4_SH(out0, out1, out2, out3, 7);
         tmp0 = PCKEV_XORI128_UB(out0, out1);
         tmp1 = PCKEV_XORI128_UB(out2, out3);
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
     }
 }
@@ -510,7 +509,7 @@  static void common_vt_8t_4w_msa(const uint8_t *src, int32_t src_stride,
         SRARI_H2_SH(out10, out32, 7);
         SAT_SH2_SH(out10, out32, 7);
         out = PCKEV_XORI128_UB(out10, out32);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         src2110 = src6554;
@@ -562,7 +561,7 @@  static void common_vt_8t_8w_msa(const uint8_t *src, int32_t src_stride,
         SAT_SH4_SH(out0_r, out1_r, out2_r, out3_r, 7);
         tmp0 = PCKEV_XORI128_UB(out0_r, out1_r);
         tmp1 = PCKEV_XORI128_UB(out2_r, out3_r);
-        ST8x4_UB(tmp0, tmp1, dst, dst_stride);
+        ST_D4(tmp0, tmp1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         src10_r = src54_r;
@@ -825,7 +824,7 @@  static void common_hv_8ht_8vt_4w_msa(const uint8_t *src, int32_t src_stride,
         SRARI_H2_SH(tmp0, tmp1, 7);
         SAT_SH2_SH(tmp0, tmp1, 7);
         out = PCKEV_XORI128_UB(tmp0, tmp1);
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         hz_out5 = hz_out9;
@@ -920,7 +919,7 @@  static void common_hv_8ht_8vt_8w_msa(const uint8_t *src, int32_t src_stride,
         SAT_SH4_SH(tmp0, tmp1, tmp2, tmp3, 7);
         vec0 = PCKEV_XORI128_UB(tmp0, tmp1);
         vec1 = PCKEV_XORI128_UB(tmp2, tmp3);
-        ST8x4_UB(vec0, vec1, dst, dst_stride);
+        ST_D4(vec0, vec1, 0, 1, 0, 1, dst, dst_stride);
         dst += (4 * dst_stride);
 
         hz_out6 = hz_out10;
@@ -1016,7 +1015,7 @@  static void common_hz_8t_and_aver_dst_4x4_msa(const uint8_t *src,
     SAT_SH2_SH(res0, res1, 7);
     res = PCKEV_XORI128_UB(res0, res1);
     res = (v16u8) __msa_aver_u_b(res, dst0);
-    ST4x4_UB(res, res, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(res, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void common_hz_8t_and_aver_dst_4x8_msa(const uint8_t *src,
@@ -1061,7 +1060,7 @@  static void common_hz_8t_and_aver_dst_4x8_msa(const uint8_t *src,
     ILVR_D2_UB(res1, res0, res3, res2, res0, res2);
     XORI_B2_128_UB(res0, res2);
     AVER_UB2_UB(res0, dst0, res2, dst1, res0, res2);
-    ST4x8_UB(res0, res2, dst, dst_stride);
+    ST_W8(res0, res2, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void common_hz_8t_and_aver_dst_4w_msa(const uint8_t *src,
@@ -1348,7 +1347,7 @@  static void common_vt_8t_and_aver_dst_4w_msa(const uint8_t *src,
         out = PCKEV_XORI128_UB(out10, out32);
         out = __msa_aver_u_b(out, dst0);
 
-        ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         src2110 = src6554;
@@ -1619,7 +1618,7 @@  static void common_hv_8ht_8vt_and_aver_dst_4w_msa(const uint8_t *src,
         SAT_SH2_SH(res0, res1, 7);
         res = PCKEV_XORI128_UB(res0, res1);
         res = (v16u8) __msa_aver_u_b(res, dst0);
-        ST4x4_UB(res, res, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(res, 0, 1, 2, 3, dst, dst_stride);
         dst += (4 * dst_stride);
 
         hz_out5 = hz_out9;
@@ -1812,7 +1811,8 @@  static void common_hz_2t_4x4_msa(const uint8_t *src, int32_t src_stride,
     DOTP_UB2_UH(vec0, vec1, filt0, filt0, vec2, vec3);
     SRARI_H2_UH(vec2, vec3, 7);
     PCKEV_B2_UB(vec2, vec2, vec3, vec3, res0, res1);
-    ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
+    ST_W2(res0, 0, 1, dst, dst_stride);
+    ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
 }
 
 static void common_hz_2t_4x8_msa(const uint8_t *src, int32_t src_stride,
@@ -1838,9 +1838,10 @@  static void common_hz_2t_4x8_msa(const uint8_t *src, int32_t src_stride,
     SRARI_H4_UH(vec4, vec5, vec6, vec7, 7);
     PCKEV_B4_SB(vec4, vec4, vec5, vec5, vec6, vec6, vec7, vec7,
                 res0, res1, res2, res3);
-    ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST4x4_UB(res2, res3, 0, 1, 0, 1, dst, dst_stride);
+    ST_W2(res0, 0, 1, dst, dst_stride);
+    ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
+    ST_W2(res2, 0, 1, dst + 4 * dst_stride, dst_stride);
+    ST_W2(res3, 0, 1, dst + 6 * dst_stride, dst_stride);
 }
 
 void ff_put_bilin_4h_msa(uint8_t *dst, ptrdiff_t dst_stride,
@@ -1877,7 +1878,7 @@  static void common_hz_2t_8x4_msa(const uint8_t *src, int32_t src_stride,
                 vec0, vec1, vec2, vec3);
     SRARI_H4_UH(vec0, vec1, vec2, vec3, 7);
     PCKEV_B2_SB(vec1, vec0, vec3, vec2, src0, src1);
-    ST8x4_UB(src0, src1, dst, dst_stride);
+    ST_D4(src0, src1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void common_hz_2t_8x8mult_msa(const uint8_t *src, int32_t src_stride,
@@ -1906,8 +1907,7 @@  static void common_hz_2t_8x8mult_msa(const uint8_t *src, int32_t src_stride,
     src += (4 * src_stride);
 
     PCKEV_B2_SB(vec1, vec0, vec3, vec2, out0, out1);
-    ST8x4_UB(out0, out1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 
     VSHF_B2_UH(src0, src0, src1, src1, mask, mask, vec0, vec1);
     VSHF_B2_UH(src2, src2, src3, src3, mask, mask, vec2, vec3);
@@ -1915,8 +1915,8 @@  static void common_hz_2t_8x8mult_msa(const uint8_t *src, int32_t src_stride,
                 vec0, vec1, vec2, vec3);
     SRARI_H4_UH(vec0, vec1, vec2, vec3, 7);
     PCKEV_B2_SB(vec1, vec0, vec3, vec2, out0, out1);
-    ST8x4_UB(out0, out1, dst, dst_stride);
-    dst += (4 * dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
+    dst += (8 * dst_stride);
 
     if (16 == height) {
         LD_SB4(src, src_stride, src0, src1, src2, src3);
@@ -1931,7 +1931,7 @@  static void common_hz_2t_8x8mult_msa(const uint8_t *src, int32_t src_stride,
         src += (4 * src_stride);
 
         PCKEV_B2_SB(vec1, vec0, vec3, vec2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 
         VSHF_B2_UH(src0, src0, src1, src1, mask, mask, vec0, vec1);
         VSHF_B2_UH(src2, src2, src3, src3, mask, mask, vec2, vec3);
@@ -1939,7 +1939,7 @@  static void common_hz_2t_8x8mult_msa(const uint8_t *src, int32_t src_stride,
                     vec0, vec1, vec2, vec3);
         SRARI_H4_UH(vec0, vec1, vec2, vec3, 7);
         PCKEV_B2_SB(vec1, vec0, vec3, vec2, out0, out1);
-        ST8x4_UB(out0, out1, dst + 4 * dst_stride, dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
     }
 }
 
@@ -2137,7 +2137,7 @@  static void common_vt_2t_4x4_msa(const uint8_t *src, int32_t src_stride,
     SRARI_H2_UH(tmp0, tmp1, 7);
     SAT_UH2_UH(tmp0, tmp1, 7);
     src2110 = __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
-    ST4x4_UB(src2110, src2110, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(src2110, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void common_vt_2t_4x8_msa(const uint8_t *src, int32_t src_stride,
@@ -2171,8 +2171,7 @@  static void common_vt_2t_4x8_msa(const uint8_t *src, int32_t src_stride,
     SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, src2110, src4332);
-    ST4x4_UB(src2110, src2110, 0, 1, 2, 3, dst, dst_stride);
-    ST4x4_UB(src4332, src4332, 0, 1, 2, 3, dst + 4 * dst_stride, dst_stride);
+    ST_W8(src2110, src4332, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
 }
 
 void ff_put_bilin_4v_msa(uint8_t *dst, ptrdiff_t dst_stride,
@@ -2209,7 +2208,7 @@  static void common_vt_2t_8x4_msa(const uint8_t *src, int32_t src_stride,
     SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-    ST8x4_UB(out0, out1, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void common_vt_2t_8x8mult_msa(const uint8_t *src, int32_t src_stride,
@@ -2243,16 +2242,15 @@  static void common_vt_2t_8x8mult_msa(const uint8_t *src, int32_t src_stride,
         SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
         SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
         PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
-        dst += (4 * dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 
         DOTP_UB4_UH(vec4, vec5, vec6, vec7, filt0, filt0, filt0, filt0,
                     tmp0, tmp1, tmp2, tmp3);
         SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
         SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
         PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
-        dst += (4 * dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
+        dst += (8 * dst_stride);
 
         src0 = src8;
     }
@@ -2514,7 +2512,8 @@  static void common_hv_2ht_2vt_4x4_msa(const uint8_t *src, int32_t src_stride,
     SRARI_H2_UH(tmp0, tmp1, 7);
     SAT_UH2_UH(tmp0, tmp1, 7);
     PCKEV_B2_UB(tmp0, tmp0, tmp1, tmp1, res0, res1);
-    ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
+    ST_W2(res0, 0, 1, dst, dst_stride);
+    ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
 }
 
 static void common_hv_2ht_2vt_4x8_msa(const uint8_t *src, int32_t src_stride,
@@ -2557,9 +2556,10 @@  static void common_hv_2ht_2vt_4x8_msa(const uint8_t *src, int32_t src_stride,
     SAT_UH4_UH(vec4, vec5, vec6, vec7, 7);
     PCKEV_B4_SB(vec4, vec4, vec5, vec5, vec6, vec6, vec7, vec7,
                 res0, res1, res2, res3);
-    ST4x4_UB(res0, res1, 0, 1, 0, 1, dst, dst_stride);
-    dst += (4 * dst_stride);
-    ST4x4_UB(res2, res3, 0, 1, 0, 1, dst, dst_stride);
+    ST_W2(res0, 0, 1, dst, dst_stride);
+    ST_W2(res1, 0, 1, dst + 2 * dst_stride, dst_stride);
+    ST_W2(res2, 0, 1, dst + 4 * dst_stride, dst_stride);
+    ST_W2(res3, 0, 1, dst + 6 * dst_stride, dst_stride);
 }
 
 void ff_put_bilin_4hv_msa(uint8_t *dst, ptrdiff_t dst_stride,
@@ -2618,7 +2618,7 @@  static void common_hv_2ht_2vt_8x4_msa(const uint8_t *src, int32_t src_stride,
     SRARI_H4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     PCKEV_B2_SB(tmp1, tmp0, tmp3, tmp2, out0, out1);
-    ST8x4_UB(out0, out1, dst, dst_stride);
+    ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 }
 
 static void common_hv_2ht_2vt_8x8mult_msa(const uint8_t *src, int32_t src_stride,
@@ -2674,8 +2674,7 @@  static void common_hv_2ht_2vt_8x8mult_msa(const uint8_t *src, int32_t src_stride
         SRARI_H2_UH(tmp3, tmp4, 7);
         SAT_UH2_UH(tmp3, tmp4, 7);
         PCKEV_B2_SB(tmp2, tmp1, tmp4, tmp3, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
-        dst += (4 * dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst, dst_stride);
 
         hz_out1 = HORIZ_2TAP_FILT_UH(src1, src1, mask, filt_hz, 7);
         vec0 = (v16u8) __msa_ilvev_b((v16i8) hz_out1, (v16i8) hz_out0);
@@ -2696,8 +2695,8 @@  static void common_hv_2ht_2vt_8x8mult_msa(const uint8_t *src, int32_t src_stride
         SRARI_H4_UH(tmp5, tmp6, tmp7, tmp8, 7);
         SAT_UH4_UH(tmp5, tmp6, tmp7, tmp8, 7);
         PCKEV_B2_SB(tmp6, tmp5, tmp8, tmp7, out0, out1);
-        ST8x4_UB(out0, out1, dst, dst_stride);
-        dst += (4 * dst_stride);
+        ST_D4(out0, out1, 0, 1, 0, 1, dst + 4 * dst_stride, dst_stride);
+        dst += (8 * dst_stride);
     }
 }
 
@@ -2842,7 +2841,7 @@  static void common_hz_2t_and_aver_dst_4x4_msa(const uint8_t *src,
     res = (v16u8) __msa_pckev_b((v16i8) vec3, (v16i8) vec2);
     res = (v16u8) __msa_aver_u_b(res, dst0);
 
-    ST4x4_UB(res, res, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(res, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void common_hz_2t_and_aver_dst_4x8_msa(const uint8_t *src,
@@ -2876,7 +2875,7 @@  static void common_hz_2t_and_aver_dst_4x8_msa(const uint8_t *src,
                 res2, res3);
     ILVR_D2_UB(res1, res0, res3, res2, res0, res2);
     AVER_UB2_UB(res0, dst0, res2, dst1, res0, res2);
-    ST4x8_UB(res0, res2, dst, dst_stride);
+    ST_W8(res0, res2, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
 }
 
 void ff_avg_bilin_4h_msa(uint8_t *dst, ptrdiff_t dst_stride,
@@ -3202,7 +3201,7 @@  static void common_vt_2t_and_aver_dst_4x4_msa(const uint8_t *src,
     out = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
     out = __msa_aver_u_b(out, dst0);
 
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void common_vt_2t_and_aver_dst_4x8_msa(const uint8_t *src,
@@ -3241,7 +3240,7 @@  static void common_vt_2t_and_aver_dst_4x8_msa(const uint8_t *src,
     SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, src2110, src4332);
     AVER_UB2_UB(src2110, dst0, src4332, dst1, src2110, src4332);
-    ST4x8_UB(src2110, src4332, dst, dst_stride);
+    ST_W8(src2110, src4332, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
 }
 
 void ff_avg_bilin_4v_msa(uint8_t *dst, ptrdiff_t dst_stride,
@@ -3620,7 +3619,7 @@  static void common_hv_2ht_2vt_and_aver_dst_4x4_msa(const uint8_t *src,
     out = (v16u8) __msa_pckev_b((v16i8) tmp1, (v16i8) tmp0);
     out = __msa_aver_u_b(out, dst0);
 
-    ST4x4_UB(out, out, 0, 1, 2, 3, dst, dst_stride);
+    ST_W4(out, 0, 1, 2, 3, dst, dst_stride);
 }
 
 static void common_hv_2ht_2vt_and_aver_dst_4x8_msa(const uint8_t *src,
@@ -3672,7 +3671,7 @@  static void common_hv_2ht_2vt_and_aver_dst_4x8_msa(const uint8_t *src,
     SAT_UH4_UH(tmp0, tmp1, tmp2, tmp3, 7);
     PCKEV_B2_UB(tmp1, tmp0, tmp3, tmp2, res0, res1);
     AVER_UB2_UB(res0, dst0, res1, dst1, res0, res1);
-    ST4x8_UB(res0, res1, dst, dst_stride);
+    ST_W8(res0, res1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
 }
 
 void ff_avg_bilin_4hv_msa(uint8_t *dst, ptrdiff_t dst_stride,
@@ -4071,14 +4070,14 @@  static void avg_width4_msa(const uint8_t *src, int32_t src_stride,
         LW4(dst + 4 * dst_stride, dst_stride, tp0, tp1, tp2, tp3);
         INSERT_W4_UB(tp0, tp1, tp2, tp3, dst1);
         AVER_UB2_UB(src0, dst0, src1, dst1, dst0, dst1);
-        ST4x8_UB(dst0, dst1, dst, dst_stride);
+        ST_W8(dst0, dst1, 0, 1, 2, 3, 0, 1, 2, 3, dst, dst_stride);
     } else if (4 == height) {
         LW4(src, src_stride, tp0, tp1, tp2, tp3);
         INSERT_W4_UB(tp0, tp1, tp2, tp3, src0);
         LW4(dst, dst_stride, tp0, tp1, tp2, tp3);
         INSERT_W4_UB(tp0, tp1, tp2, tp3, dst0);
         dst0 = __msa_aver_u_b(src0, dst0);
-        ST4x4_UB(dst0, dst0, 0, 1, 2, 3, dst, dst_stride);
+        ST_W4(dst0, 0, 1, 2, 3, dst, dst_stride);
     }
 }
 
@@ -4109,7 +4108,7 @@  static void avg_width8_msa(const uint8_t *src, int32_t src_stride,
             INSERT_D2_UB(tp6, tp7, dst3);
             AVER_UB4_UB(src0, dst0, src1, dst1, src2, dst2, src3, dst3, dst0,
                         dst1, dst2, dst3);
-            ST8x8_UB(dst0, dst1, dst2, dst3, dst, dst_stride);
+            ST_D8(dst0, dst1, dst2, dst3, 0, 1, 0, 1, 0, 1, 0, 1, dst, dst_stride);
             dst += 8 * dst_stride;
         }
     } else if (4 == height) {
@@ -4120,7 +4119,7 @@  static void avg_width8_msa(const uint8_t *src, int32_t src_stride,
         INSERT_D2_UB(tp0, tp1, dst0);
         INSERT_D2_UB(tp2, tp3, dst1);
         AVER_UB2_UB(src0, dst0, src1, dst1, dst0, dst1);
-        ST8x4_UB(dst0, dst1, dst, dst_stride);
+        ST_D4(dst0, dst1, 0, 1, 0, 1, dst, dst_stride);
     }
 }
 
diff --git a/libavutil/mips/generic_macros_msa.h b/libavutil/mips/generic_macros_msa.h
index a377428..9ac0583 100644
--- a/libavutil/mips/generic_macros_msa.h
+++ b/libavutil/mips/generic_macros_msa.h
@@ -349,19 +349,6 @@ 
 }
 #define LD_SH16(...) LD_V16(v8i16, __VA_ARGS__)
 
-/* Description : Load as 4x4 block of signed halfword elements from 1D source
-                 data into 4 vectors (Each vector with 4 signed halfwords)
-   Arguments   : Inputs  - psrc
-                 Outputs - out0, out1, out2, out3
-*/
-#define LD4x4_SH(psrc, out0, out1, out2, out3)                \
-{                                                             \
-    out0 = LD_SH(psrc);                                       \
-    out2 = LD_SH(psrc + 8);                                   \
-    out1 = (v8i16) __msa_ilvl_d((v2i64) out0, (v2i64) out0);  \
-    out3 = (v8i16) __msa_ilvl_d((v2i64) out2, (v2i64) out2);  \
-}
-
 /* Description : Store vectors with stride
    Arguments   : Inputs  - in0, in1, stride
                  Outputs - pdst    (destination pointer to store to)
@@ -405,198 +392,127 @@ 
 #define ST_SH8(...) ST_V8(v8i16, __VA_ARGS__)
 #define ST_SW8(...) ST_V8(v4i32, __VA_ARGS__)
 
-/* Description : Store as 2x4 byte block to destination memory from input vector
-   Arguments   : Inputs  - in, stidx, pdst, stride
-                 Return Type - unsigned byte
-   Details     : Index stidx halfword element from 'in' vector is copied and
-                 stored on first line
-                 Index stidx+1 halfword element from 'in' vector is copied and
-                 stored on second line
-                 Index stidx+2 halfword element from 'in' vector is copied and
-                 stored on third line
-                 Index stidx+3 halfword element from 'in' vector is copied and
-                 stored on fourth line
-*/
-#define ST2x4_UB(in, stidx, pdst, stride)              \
-{                                                      \
-    uint16_t out0_m, out1_m, out2_m, out3_m;           \
-    uint8_t *pblk_2x4_m = (uint8_t *) (pdst);          \
-                                                       \
-    out0_m = __msa_copy_u_h((v8i16) in, (stidx));      \
-    out1_m = __msa_copy_u_h((v8i16) in, (stidx + 1));  \
-    out2_m = __msa_copy_u_h((v8i16) in, (stidx + 2));  \
-    out3_m = __msa_copy_u_h((v8i16) in, (stidx + 3));  \
-                                                       \
-    SH(out0_m, pblk_2x4_m);                            \
-    SH(out1_m, pblk_2x4_m + stride);                   \
-    SH(out2_m, pblk_2x4_m + 2 * stride);               \
-    SH(out3_m, pblk_2x4_m + 3 * stride);               \
-}
-
-/* Description : Store as 4x2 byte block to destination memory from input vector
-   Arguments   : Inputs  - in, pdst, stride
-                 Return Type - unsigned byte
-   Details     : Index 0 word element from input vector is copied and stored
-                 on first line
-                 Index 1 word element from input vector is copied and stored
-                 on second line
-*/
-#define ST4x2_UB(in, pdst, stride)             \
-{                                              \
-    uint32_t out0_m, out1_m;                   \
-    uint8_t *pblk_4x2_m = (uint8_t *) (pdst);  \
-                                               \
-    out0_m = __msa_copy_u_w((v4i32) in, 0);    \
-    out1_m = __msa_copy_u_w((v4i32) in, 1);    \
-                                               \
-    SW(out0_m, pblk_4x2_m);                    \
-    SW(out1_m, pblk_4x2_m + stride);           \
-}
-
-/* Description : Store as 4x4 byte block to destination memory from input vector
-   Arguments   : Inputs  - in0, in1, pdst, stride
-                 Return Type - unsigned byte
-   Details     : Idx0 word element from input vector 'in0' is copied and stored
-                 on first line
-                 Idx1 word element from input vector 'in0' is copied and stored
-                 on second line
-                 Idx2 word element from input vector 'in1' is copied and stored
-                 on third line
-                 Idx3 word element from input vector 'in1' is copied and stored
-                 on fourth line
-*/
-#define ST4x4_UB(in0, in1, idx0, idx1, idx2, idx3, pdst, stride)  \
-{                                                                 \
-    uint32_t out0_m, out1_m, out2_m, out3_m;                      \
-    uint8_t *pblk_4x4_m = (uint8_t *) (pdst);                     \
-                                                                  \
-    out0_m = __msa_copy_u_w((v4i32) in0, idx0);                   \
-    out1_m = __msa_copy_u_w((v4i32) in0, idx1);                   \
-    out2_m = __msa_copy_u_w((v4i32) in1, idx2);                   \
-    out3_m = __msa_copy_u_w((v4i32) in1, idx3);                   \
-                                                                  \
-    SW4(out0_m, out1_m, out2_m, out3_m, pblk_4x4_m, stride);      \
+/* Description : Store half word elements of vector with stride
+ * Arguments   : Inputs  - in   source vector
+ *                       - pdst    (destination pointer to store to)
+ *                       - stride
+ * Details     : Stores half word 'idx0' from 'in' to (pdst)
+ *               Stores half word 'idx1' from 'in' to (pdst + stride)
+ *               Similar for other elements
+ */
+#define ST_H1(in, idx, pdst)                             \
+{                                                        \
+    uint16_t out0_m;                                     \
+    out0_m = __msa_copy_u_h((v8i16) in, idx);            \
+    SH(out0_m, (pdst));                                  \
 }
-#define ST4x8_UB(in0, in1, pdst, stride)                            \
-{                                                                   \
-    uint8_t *pblk_4x8 = (uint8_t *) (pdst);                         \
-                                                                    \
-    ST4x4_UB(in0, in0, 0, 1, 2, 3, pblk_4x8, stride);               \
-    ST4x4_UB(in1, in1, 0, 1, 2, 3, pblk_4x8 + 4 * stride, stride);  \
+#define ST_H2(in, idx0, idx1, pdst, stride)              \
+{                                                        \
+    uint16_t out0_m, out1_m;                             \
+    out0_m = __msa_copy_u_h((v8i16) in, idx0);           \
+    out1_m = __msa_copy_u_h((v8i16) in, idx1);           \
+    SH(out0_m, (pdst));                                  \
+    SH(out1_m, (pdst) + stride);                         \
+}
+#define ST_H4(in, idx0, idx1, idx2, idx3, pdst, stride)  \
+{                                                        \
+    uint16_t out0_m, out1_m, out2_m, out3_m;             \
+    out0_m = __msa_copy_u_h((v8i16) in, idx0);           \
+    out1_m = __msa_copy_u_h((v8i16) in, idx1);           \
+    out2_m = __msa_copy_u_h((v8i16) in, idx2);           \
+    out3_m = __msa_copy_u_h((v8i16) in, idx3);           \
+    SH(out0_m, (pdst));                                  \
+    SH(out1_m, (pdst) + stride);                         \
+    SH(out2_m, (pdst) + 2 * stride);                     \
+    SH(out3_m, (pdst) + 3 * stride);                     \
+}
+#define ST_H8(in, idx0, idx1, idx2, idx3, idx4, idx5,            \
+              idx6, idx7, pdst, stride)                          \
+{                                                                \
+    ST_H4(in, idx0, idx1, idx2, idx3, pdst, stride)              \
+    ST_H4(in, idx4, idx5, idx6, idx7, (pdst) + 4*stride, stride) \
 }
 
-/* Description : Store as 6x4 byte block to destination memory from input
-                 vectors
-   Arguments   : Inputs  - in0, in1, pdst, stride
-                 Return Type - unsigned byte
-   Details     : Index 0 word element from input vector 'in0' is copied and
-                 stored on first line followed by index 2 halfword element
-                 Index 2 word element from input vector 'in0' is copied and
-                 stored on second line followed by index 2 halfword element
-                 Index 0 word element from input vector 'in1' is copied and
-                 stored on third line followed by index 2 halfword element
-                 Index 2 word element from input vector 'in1' is copied and
-                 stored on fourth line followed by index 2 halfword element
-*/
-#define ST6x4_UB(in0, in1, pdst, stride)       \
+/* Description : Store word elements of vector with stride
+ * Arguments   : Inputs  - in   source vector
+ *                       - pdst    (destination pointer to store to)
+ *                       - stride
+ * Details     : Stores word 'idx0' from 'in' to (pdst)
+ *               Stores word 'idx1' from 'in' to (pdst + stride)
+ *               Similar for other elements
+ */
+#define ST_W1(in, idx, pdst)                             \
+{                                                        \
+    uint32_t out0_m;                                     \
+    out0_m = __msa_copy_u_w((v4i32) in, idx);            \
+    SW(out0_m, (pdst));                                  \
+}
+#define ST_W2(in, idx0, idx1, pdst, stride)              \
+{                                                        \
+    uint32_t out0_m, out1_m;                             \
+    out0_m = __msa_copy_u_w((v4i32) in, idx0);           \
+    out1_m = __msa_copy_u_w((v4i32) in, idx1);           \
+    SW(out0_m, (pdst));                                  \
+    SW(out1_m, (pdst) + stride);                         \
+}
+#define ST_W4(in, idx0, idx1, idx2, idx3, pdst, stride)  \
+{                                                        \
+    uint32_t out0_m, out1_m, out2_m, out3_m;             \
+    out0_m = __msa_copy_u_w((v4i32) in, idx0);           \
+    out1_m = __msa_copy_u_w((v4i32) in, idx1);           \
+    out2_m = __msa_copy_u_w((v4i32) in, idx2);           \
+    out3_m = __msa_copy_u_w((v4i32) in, idx3);           \
+    SW(out0_m, (pdst));                                  \
+    SW(out1_m, (pdst) + stride);                         \
+    SW(out2_m, (pdst) + 2*stride);                       \
+    SW(out3_m, (pdst) + 3*stride);                       \
+}
+#define ST_W8(in0, in1, idx0, idx1, idx2, idx3,                 \
+              idx4, idx5, idx6, idx7, pdst, stride)             \
+{                                                               \
+    ST_W4(in0, idx0, idx1, idx2, idx3, pdst, stride)            \
+    ST_W4(in1, idx4, idx5, idx6, idx7, pdst + 4*stride, stride) \
+}
+
+/* Description : Store double word elements of vector with stride
+ * Arguments   : Inputs  - in   source vector
+ *                       - pdst    (destination pointer to store to)
+ *                       - stride
+ * Details     : Stores double word 'idx0' from 'in' to (pdst)
+ *               Stores double word 'idx1' from 'in' to (pdst + stride)
+ *               Similar for other elements
+ */
+#define ST_D1(in, idx, pdst)                   \
 {                                              \
-    uint32_t out0_m, out1_m, out2_m, out3_m;   \
-    uint16_t out4_m, out5_m, out6_m, out7_m;   \
-    uint8_t *pblk_6x4_m = (uint8_t *) (pdst);  \
-                                               \
-    out0_m = __msa_copy_u_w((v4i32) in0, 0);   \
-    out1_m = __msa_copy_u_w((v4i32) in0, 2);   \
-    out2_m = __msa_copy_u_w((v4i32) in1, 0);   \
-    out3_m = __msa_copy_u_w((v4i32) in1, 2);   \
-                                               \
-    out4_m = __msa_copy_u_h((v8i16) in0, 2);   \
-    out5_m = __msa_copy_u_h((v8i16) in0, 6);   \
-    out6_m = __msa_copy_u_h((v8i16) in1, 2);   \
-    out7_m = __msa_copy_u_h((v8i16) in1, 6);   \
-                                               \
-    SW(out0_m, pblk_6x4_m);                    \
-    SH(out4_m, (pblk_6x4_m + 4));              \
-    pblk_6x4_m += stride;                      \
-    SW(out1_m, pblk_6x4_m);                    \
-    SH(out5_m, (pblk_6x4_m + 4));              \
-    pblk_6x4_m += stride;                      \
-    SW(out2_m, pblk_6x4_m);                    \
-    SH(out6_m, (pblk_6x4_m + 4));              \
-    pblk_6x4_m += stride;                      \
-    SW(out3_m, pblk_6x4_m);                    \
-    SH(out7_m, (pblk_6x4_m + 4));              \
-}
-
-/* Description : Store as 8x1 byte block to destination memory from input vector
-   Arguments   : Inputs  - in, pdst
-   Details     : Index 0 double word element from input vector 'in' is copied
-                 and stored to destination memory at (pdst)
-*/
-#define ST8x1_UB(in, pdst)                   \
-{                                            \
-    uint64_t out0_m;                         \
-    out0_m = __msa_copy_u_d((v2i64) in, 0);  \
-    SD(out0_m, pdst);                        \
-}
-
-/* Description : Store as 8x2 byte block to destination memory from input vector
-   Arguments   : Inputs  - in, pdst, stride
-   Details     : Index 0 double word element from input vector 'in' is copied
-                 and stored to destination memory at (pdst)
-                 Index 1 double word element from input vector 'in' is copied
-                 and stored to destination memory at (pdst + stride)
-*/
-#define ST8x2_UB(in, pdst, stride)             \
+    uint64_t out0_m;                           \
+    out0_m = __msa_copy_u_d((v2i64) in, idx);  \
+    SD(out0_m, (pdst));                        \
+}
+#define ST_D2(in, idx0, idx1, pdst, stride)    \
 {                                              \
     uint64_t out0_m, out1_m;                   \
-    uint8_t *pblk_8x2_m = (uint8_t *) (pdst);  \
-                                               \
-    out0_m = __msa_copy_u_d((v2i64) in, 0);    \
-    out1_m = __msa_copy_u_d((v2i64) in, 1);    \
-                                               \
-    SD(out0_m, pblk_8x2_m);                    \
-    SD(out1_m, pblk_8x2_m + stride);           \
-}
-
-/* Description : Store as 8x4 byte block to destination memory from input
-                 vectors
-   Arguments   : Inputs  - in0, in1, pdst, stride
-   Details     : Index 0 double word element from input vector 'in0' is copied
-                 and stored to destination memory at (pblk_8x4_m)
-                 Index 1 double word element from input vector 'in0' is copied
-                 and stored to destination memory at (pblk_8x4_m + stride)
-                 Index 0 double word element from input vector 'in1' is copied
-                 and stored to destination memory at (pblk_8x4_m + 2 * stride)
-                 Index 1 double word element from input vector 'in1' is copied
-                 and stored to destination memory at (pblk_8x4_m + 3 * stride)
-*/
-#define ST8x4_UB(in0, in1, pdst, stride)                      \
-{                                                             \
-    uint64_t out0_m, out1_m, out2_m, out3_m;                  \
-    uint8_t *pblk_8x4_m = (uint8_t *) (pdst);                 \
-                                                              \
-    out0_m = __msa_copy_u_d((v2i64) in0, 0);                  \
-    out1_m = __msa_copy_u_d((v2i64) in0, 1);                  \
-    out2_m = __msa_copy_u_d((v2i64) in1, 0);                  \
-    out3_m = __msa_copy_u_d((v2i64) in1, 1);                  \
-                                                              \
-    SD4(out0_m, out1_m, out2_m, out3_m, pblk_8x4_m, stride);  \
+    out0_m = __msa_copy_u_d((v2i64) in, idx0); \
+    out1_m = __msa_copy_u_d((v2i64) in, idx1); \
+    SD(out0_m, (pdst));                        \
+    SD(out1_m, (pdst) + stride);               \
 }
-#define ST8x8_UB(in0, in1, in2, in3, pdst, stride)        \
-{                                                         \
-    uint8_t *pblk_8x8_m = (uint8_t *) (pdst);             \
-                                                          \
-    ST8x4_UB(in0, in1, pblk_8x8_m, stride);               \
-    ST8x4_UB(in2, in3, pblk_8x8_m + 4 * stride, stride);  \
-}
-#define ST12x4_UB(in0, in1, in2, pdst, stride)                \
+#define ST_D4(in0, in1, idx0, idx1, idx2, idx3, pdst, stride) \
 {                                                             \
-    uint8_t *pblk_12x4_m = (uint8_t *) (pdst);                \
-                                                              \
-    /* left 8x4 */                                            \
-    ST8x4_UB(in0, in1, pblk_12x4_m, stride);                  \
-    /* right 4x4 */                                           \
-    ST4x4_UB(in2, in2, 0, 1, 2, 3, pblk_12x4_m + 8, stride);  \
+    uint64_t out0_m, out1_m, out2_m, out3_m;                  \
+    out0_m = __msa_copy_u_d((v2i64) in0, idx0);               \
+    out1_m = __msa_copy_u_d((v2i64) in0, idx1);               \
+    out2_m = __msa_copy_u_d((v2i64) in1, idx2);               \
+    out3_m = __msa_copy_u_d((v2i64) in1, idx3);               \
+    SD(out0_m, (pdst));                                       \
+    SD(out1_m, (pdst) + stride);                              \
+    SD(out2_m, (pdst) + 2 * stride);                          \
+    SD(out3_m, (pdst) + 3 * stride);                          \
+}
+#define ST_D8(in0, in1, in2, in3, idx0, idx1, idx2, idx3,              \
+              idx4, idx5, idx6, idx7, pdst, stride)                    \
+{                                                                      \
+    ST_D4(in0, in1, idx0, idx1, idx2, idx3, pdst, stride)              \
+    ST_D4(in2, in3, idx4, idx5, idx6, idx7, pdst + 4 * stride, stride) \
 }
 
 /* Description : Store as 12x8 byte block to destination memory from
@@ -2926,7 +2842,7 @@ 
     tmp0_m = PCKEV_XORI128_UB(in0, in1);                      \
     tmp1_m = PCKEV_XORI128_UB(in2, in3);                      \
     AVER_UB2_UB(tmp0_m, dst0, tmp1_m, dst1, tmp0_m, tmp1_m);  \
-    ST8x4_UB(tmp0_m, tmp1_m, pdst_m, stride);                 \
+    ST_D4(tmp0_m, tmp1_m, 0, 1, 0, 1, pdst_m, stride);        \
 }
 
 /* Description : Pack even byte elements, extract 0 & 2 index words from pair