diff mbox series

[FFmpeg-devel] aacdec: restore arm32 dequantization optimizations

Message ID 20240511131428.1733729-1-dev@lynne.ee
State Accepted
Commit baf8651d563bd42b27738e5215c0d822638c0a9b
Headers show
Series [FFmpeg-devel] aacdec: restore arm32 dequantization optimizations | expand

Commit Message

Lynne May 11, 2024, 1:14 p.m. UTC
Unintentionally removed as part of 03cf10164578aed33f4d0cb5b69d63669c01a538.
Untested, but its assumed that unlike most of the old ARM code,
this one was still working.
---
 libavcodec/aac/aacdec_float.c | 5 +++++
 1 file changed, 5 insertions(+)

Comments

Martin Storsjö May 13, 2024, 10:13 a.m. UTC | #1
On Sat, 11 May 2024, Lynne via ffmpeg-devel wrote:

> Unintentionally removed as part of 03cf10164578aed33f4d0cb5b69d63669c01a538.
> Untested, but its assumed that unlike most of the old ARM code,
> this one was still working.
> ---
> libavcodec/aac/aacdec_float.c | 5 +++++
> 1 file changed, 5 insertions(+)

LGTM, thanks. Tested and it does seem to pass fate successfully on arm32 
builds.

// Martin
Lynne May 13, 2024, 11:03 a.m. UTC | #2
On 13/05/2024 12:13, Martin Storsjö wrote:
> On Sat, 11 May 2024, Lynne via ffmpeg-devel wrote:
> 
>> Unintentionally removed as part of 
>> 03cf10164578aed33f4d0cb5b69d63669c01a538.
>> Untested, but its assumed that unlike most of the old ARM code,
>> this one was still working.
>> ---
>> libavcodec/aac/aacdec_float.c | 5 +++++
>> 1 file changed, 5 insertions(+)
> 
> LGTM, thanks. Tested and it does seem to pass fate successfully on arm32 
> builds.
> 
> // Martin
> 

Thanks, pushed.
diff mbox series

Patch

diff --git a/libavcodec/aac/aacdec_float.c b/libavcodec/aac/aacdec_float.c
index 885d824fa7..03ec264c50 100644
--- a/libavcodec/aac/aacdec_float.c
+++ b/libavcodec/aac/aacdec_float.c
@@ -79,6 +79,11 @@  static const float cce_scale[] = {
 #include "aacdec_tab.h"
 #include "libavutil/intfloat.h"
 
+#include "config.h"
+#if ARCH_ARM
+#include "libavcodec/arm/aac.h"
+#endif
+
 #ifndef VMUL2
 static inline float *VMUL2(float *dst, const float *v, unsigned idx,
                            const float *scale)