From 64821f5a7cc85f8d7cfc69b5caaa8d766faedc2e Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Cl=C3=A9ment=20B=C5=93sch?= <u@pkh.me>
Date: Fri, 22 Nov 2013 20:16:09 +0100
Subject: [PATCH] avcodec/vp9dsp: add DC only versions for idct/idct.
before:
./ffmpeg -v 0 -nostats -i ~/samples/vp9/etv.webm -f null - 16.29s user 0.02s system 99% cpu 16.323 total
./ffmpeg -v 0 -nostats -i ~/samples/vp9/etv.webm -f null - 16.32s user 0.01s system 99% cpu 16.351 total
./ffmpeg -v 0 -nostats -i ~/samples/vp9/etv.webm -f null - 16.27s user 0.05s system 99% cpu 16.335 total
after:
./ffmpeg -v 0 -nostats -i ~/samples/vp9/etv.webm -f null - 15.22s user 0.03s system 99% cpu 15.257 total
./ffmpeg -v 0 -nostats -i ~/samples/vp9/etv.webm -f null - 15.20s user 0.02s system 99% cpu 15.237 total
./ffmpeg -v 0 -nostats -i ~/samples/vp9/etv.webm -f null - 15.19s user 0.02s system 99% cpu 15.227 total
---
libavcodec/vp9dsp.c | 22 ++++++++++++++++++++++
1 file changed, 22 insertions(+)
diff --git a/libavcodec/vp9dsp.c b/libavcodec/vp9dsp.c
index 6159616fda7..2b111ad1a13 100644
--- a/libavcodec/vp9dsp.c
+++ b/libavcodec/vp9dsp.c
@@ -853,6 +853,12 @@ static av_cold void vp9dsp_intrapred_init(VP9DSPContext *dsp)
#undef init_intra_pred
}
+#define has_dconly_idct_idct 1
+#define has_dconly_iadst_idct 0
+#define has_dconly_idct_iadst 0
+#define has_dconly_iadst_iadst 0
+#define has_dconly_iwht_iwht 0
+
#define itxfm_wrapper(type_a, type_b, sz, bits) \
static void type_a##_##type_b##_##sz##x##sz##_add_c(uint8_t *dst, \
ptrdiff_t stride, \
@@ -860,6 +866,22 @@ static void type_a##_##type_b##_##sz##x##sz##_add_c(uint8_t *dst, \
{ \
int i, j; \
int16_t tmp[sz * sz], out[sz]; \
+\
+ if (has_dconly_##type_a##_##type_b && eob == 1) { \
+ const int t = (((block[0] * 11585 + (1 << 13)) >> 14) \
+ * 11585 + (1 << 13)) >> 14; \
+ block[0] = 0; \
+ for (i = 0; i < sz; i++) { \
+ for (j = 0; j < sz; j++) \
+ dst[j * stride] = av_clip_uint8(dst[j * stride] + \
+ (bits ? \
+ (t + (1 << (bits - 1))) >> bits : \
+ t)); \
+ dst++; \
+ } \
+ return; \
+ } \
+\
for (i = 0; i < sz; i++) \
type_a##sz##_1d(block + i, sz, tmp + i * sz, 0); \
memset(block, 0, sz * sz * sizeof(*block)); \
--
GitLab