PR #23107 opened by Zhao Zhili (quink)
URL: https://code.ffmpeg.org/FFmpeg/FFmpeg/pulls/23107
Patch URL: https://code.ffmpeg.org/FFmpeg/FFmpeg/pulls/23107.patch
vc1_loop_filter() is only reached through the six C wrappers. Clang 14
keeps it out of line with plain static inline, adding a 224-byte stack
frame before the tiny bestcase path on rpi 5. gcc 12 already inlines
it.
rpi 5 clang 14:
before after
vc1_v_loop_filter4_bestcase_c 27.2 8.3 (3.3x)
vc1_h_loop_filter4_bestcase_c 26.4 10.2 (2.6x)
vc1_v_loop_filter8_bestcase_c 32.5 20.3 (1.6x)
vc1_h_loop_filter8_bestcase_c 31.7 19.5 (1.6x)
vc1_v_loop_filter16_bestcase_c 42.1 33.2 (1.3x)
vc1_h_loop_filter16_bestcase_c 41.6 25.3 (1.6x)
Signed-off-by: Zhao Zhili <[email protected]>
>From 529f28fe00eef5d3eafe63a9ea5b1c4cac4c6362 Mon Sep 17 00:00:00 2001
From: Zhao Zhili <[email protected]>
Date: Fri, 15 May 2026 21:56:53 +0800
Subject: [PATCH] avcodec/vc1dsp: always inline vc1_loop_filter
vc1_loop_filter() is only reached through the six C wrappers. Clang 14
keeps it out of line with plain static inline, adding a 224-byte stack
frame before the tiny bestcase path on rpi 5. gcc 12 already inlines
it.
rpi 5 clang 14:
before after
vc1_v_loop_filter4_bestcase_c 27.2 8.3 (3.3x)
vc1_h_loop_filter4_bestcase_c 26.4 10.2 (2.6x)
vc1_v_loop_filter8_bestcase_c 32.5 20.3 (1.6x)
vc1_h_loop_filter8_bestcase_c 31.7 19.5 (1.6x)
vc1_v_loop_filter16_bestcase_c 42.1 33.2 (1.3x)
vc1_h_loop_filter16_bestcase_c 41.6 25.3 (1.6x)
Signed-off-by: Zhao Zhili <[email protected]>
---
libavcodec/vc1dsp.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/libavcodec/vc1dsp.c b/libavcodec/vc1dsp.c
index 86b81dfc04..dd143b114c 100644
--- a/libavcodec/vc1dsp.c
+++ b/libavcodec/vc1dsp.c
@@ -196,8 +196,8 @@ static av_always_inline int vc1_filter_line(uint8_t *src,
ptrdiff_t stride, int
* @param pq block quantizer
* @see 8.6
*/
-static inline void vc1_loop_filter(uint8_t *src, ptrdiff_t step, ptrdiff_t
stride,
- int len, int pq)
+static av_always_inline void vc1_loop_filter(uint8_t *src, ptrdiff_t step,
+ ptrdiff_t stride, int len, int pq)
{
int i;
int filt3;
--
2.52.0
_______________________________________________
ffmpeg-devel mailing list -- [email protected]
To unsubscribe send an email to [email protected]