Changing details as following:
1. Modified the parameters order of SLDI_Bn. The previous order of
parameters is difficult to understand.
2. Remove the redundant macro SLDI_Bn_0 and use SLDI_Bn instead.
---
libavcodec/mips/h264dsp_msa.c | 9 ++--
libavcodec/mips/h264qpel_msa.c | 64 +
Changing details as following:
1. The previous order of parameters are irregular and difficult to
understand. Adjust the order of the parameters according to the
rule: (RTYPE, input registers, input mask/input index/..., output registers).
Most of the existing msa macros follow the rule.
2
Changing details as following:
1. Refine CLIP_SH, results are in placed to input vectors.
2. Reimplement the macro CLIP_SH/Wn_0_255. The new macro is more
efficient than before.
3. Remove CLIP_SH/Wn_0_255_MAX_SATU. CLIP_SH/Wn_0_255_MAX_SATU and
CLIP_SH/Wn_0_255 have the same function. It is n
Changing details as following:
1. Remove the local variable out_m in CLIP_SH. Results are assigned
to input vector, reduced the data replication.
2. Reimplement the macro CLIP_SH/Wn_0_255. The VP8 decoding performance
has improved by 1.1%(7.03x to 7.11x, tested on loongson 3A4000).
3. Remove
Changing details as following:
1. Remove the local variable 'out_m' in 'CLIP_SH' and store the result in
source vector.
2. Refine the implementation of macro 'CLIP_SH_0_255' and 'CLIP_SW_0_255'.
Performance of VP8 decoding has speed up about 1.1%(from 7.03x to 7.11x).
3. Remove redundant macr
Changing details as following:
1. Remove the local variable 'out_m' in 'CLIP_SH' and store the result in
source vector.
2. Refine the implementation of macro 'CLIP_SH_0_255' and 'CLIP_SW_0_255'.
Performance of VP8 decoding has speed up about 1.1%(from 7.03x to 7.11x).
Performance of H264 d
Use the macros of ADD8 to replace continuous addition operations.
---
libavcodec/mips/vp3dsp_idct_msa.c | 80 -
libavutil/mips/generic_macros_msa.h | 6 +++
2 files changed, 22 insertions(+), 64 deletions(-)
diff --git a/libavcodec/mips/vp3dsp_idct_msa.c
b/
.c b/libavcodec/mips/vc1dsp_msa.c
new file mode 100644
index 000..1619ea4
--- /dev/null
+++ b/libavcodec/mips/vc1dsp_msa.c
@@ -0,0 +1,483 @@
+/*
+ * Loongson SIMD optimized vc1dsp
+ *
+ * Copyright (c) 2019 Loongson Technology Corporation Limited
+ *gxw
+ *
+ * This file is
Change the stride argument to ptrdiff_t in the following functions:
ff_put_no_rnd_vc1_chroma_mc8_mmi, ff_put_no_rnd_vc1_chroma_mc4_mmi,
ff_avg_no_rnd_vc1_chroma_mc8_mmi, ff_avg_no_rnd_vc1_chroma_mc4_mmi.
---
libavcodec/mips/vc1dsp_mips.h | 8
libavcodec/mips/vc1dsp_mmi.c | 8
2
>>+TRANSPOSE4x4_SW_SW(in_l0, in_l1, in_l2, in_l3, t_l1, t_l2, t_l3,
t_l4);
>>+TRANSPOSE4x4_SW_SW(in_r4, in_r5, in_r6, in_r7, in_l0, in_l1, in_l2,
in_l3);
>>+TRANSPOSE4x4_SW_SW(in_l4, in_l5, in_l6, in_l7, in_l4, in_l5, in_l6,
in_l7);
>>+in_r4 = t_l1, in_r5 = t_l2,
.c b/libavcodec/mips/vc1dsp_msa.c
new file mode 100644
index 000..6e588e8
--- /dev/null
+++ b/libavcodec/mips/vc1dsp_msa.c
@@ -0,0 +1,461 @@
+/*
+ * Loongson SIMD optimized vc1dsp
+ *
+ * Copyright (c) 2019 Loongson Technology Corporation Limited
+ *gxw
+ *
+ * This file is
The AV_INPUT_BUFFER_PADDING_SIZE has been increased to 64, but the value is
still 32
in function ff_hevc_sao_edge_filter_8_msa. So, Modify the corresponding value
to 64.
Fate tests passed.
---
libavcodec/mips/hevc_lpf_sao_msa.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/
The AV_INPUT_BUFFER_PADDING_SIZE has been increased to 64, but the value is
still 32
in function ff_hevc_sao_edge_filter_8_msa. So, use AV_INPUT_BUFFER_PADDING_SIZE
directly.
Fate tests passed.
---
libavcodec/mips/hevc_lpf_sao_msa.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --g
The AV_INPUT_BUFFER_PADDING_SIZE has been increased to 64, but the value is
still 32
in function ff_hevc_sao_edge_filter_8_msa. So, use AV_INPUT_BUFFER_PADDING_SIZE
directly.
Also, use MAX_PB_SIZE directly instead of 64. Fate tests passed.
---
libavcodec/mips/hevc_lpf_sao_msa.c | 2 +-
1 file ch
-$(CONFIG_H264QPEL) += mips/h264qpel_msa.o
diff --git a/libavcodec/mips/vp3dsp_idct_msa.c
b/libavcodec/mips/vp3dsp_idct_msa.c
new file mode 100644
index 000..5427ac5
--- /dev/null
+++ b/libavcodec/mips/vp3dsp_idct_msa.c
@@ -0,0 +1,662 @@
+/*
+ * Copyright (c) 2018 gxw
+ *
+ * This
+++ b/libavcodec/mips/vp3dsp_idct_mmi.c
@@ -0,0 +1,769 @@
+/*
+ * Copyright (c) 2018 gxw
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software
diff --git a/libavcodec/mips/vp9_mc_mmi.c b/libavcodec/mips/vp9_mc_mmi.c
new file mode 100644
index 000..145bbff
--- /dev/null
+++ b/libavcodec/mips/vp9_mc_mmi.c
@@ -0,0 +1,680 @@
+/*
+ * Copyright (c) 2019 gxw
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can
> 在 2019年2月21日，上午9:55，Shiyou Yin 写道：
>
>> -Original Message-
>> From: ffmpeg-devel-boun...@ffmpeg.org
>> <mailto:ffmpeg-devel-boun...@ffmpeg.org>
>> [mailto:ffmpeg-devel-boun...@ffmpeg.org
>> <mailto:ffmpeg-devel-boun...@ffmpeg.org>
-$(CONFIG_VP9_DECODER)+= mips/vp9_mc_mmi.o
diff --git a/libavcodec/mips/vp9_mc_mmi.c b/libavcodec/mips/vp9_mc_mmi.c
new file mode 100644
index 000..58a920b
--- /dev/null
+++ b/libavcodec/mips/vp9_mc_mmi.c
@@ -0,0 +1,692 @@
+/*
+ * Copyright (c) 2019 gxw
+ *
+ * This file is part of FFmpeg
-$(CONFIG_VP9_DECODER)+= mips/vp9_mc_mmi.o
diff --git a/libavcodec/mips/vp9_mc_mmi.c b/libavcodec/mips/vp9_mc_mmi.c
new file mode 100644
index 000..e7a8387
--- /dev/null
+++ b/libavcodec/mips/vp9_mc_mmi.c
@@ -0,0 +1,628 @@
+/*
+ * Copyright (c) 2019 gxw
+ *
+ * This file is part of FFmpeg
> 在 2019年2月24日，上午10:55，Shiyou Yin 写道：
>
>
>
>> -Original Message-
>> From: ffmpeg-devel-boun...@ffmpeg.org
>> <mailto:ffmpeg-devel-boun...@ffmpeg.org>
>> [mailto:ffmpeg-devel-boun...@ffmpeg.org
>> <mailto:ffmpeg-devel-boun
21 matches
Mail list logo
