[FFmpeg-devel] [PATCH 1/3] mips: optimization for float aac decoder (core module)
Babic, Nedeljko
nbabic at mips.com
Wed Jan 16 15:56:26 CET 2013
It doesn't need to be global.
This is a leftover from development.
I will remove it, recreate this patch and resubmit it.
-Nedeljko
________________________________________
From: ffmpeg-devel-bounces at ffmpeg.org [ffmpeg-devel-bounces at ffmpeg.org] on behalf of Michael Niedermayer [michaelni at gmx.at]
Sent: Wednesday, January 16, 2013 13:02
To: FFmpeg development discussions and patches
Cc: Vulin, Mirjana (c); Lukac, Zeljko; Mirjana Vulin
Subject: Re: [FFmpeg-devel] [PATCH 1/3] mips: optimization for float aac decoder (core module)
On Wed, Jan 09, 2013 at 10:45:32AM +0100, Nedeljko Babic wrote:
> From: Mirjana Vulin <mirjana.vulin at rt-rk.com>
>
> Change-Id: I20673f7528108d4b607a25194dc4c7cf490ff1e0
> Signed-off-by: Mirjana Vulin <mvulin at mips.com>
> ---
> libavcodec/aac.h | 18 +-
> libavcodec/aacdec.c | 39 ++-
> libavcodec/mips/Makefile | 1 +
> libavcodec/mips/aacdec_mips.c | 830 +++++++++++++++++++++++++++++++++++++++
> libavcodec/mips/aacdec_mips.h | 251 ++++++++++++
> libavcodec/mips/dsputil_mips.c | 84 ++++-
> libavutil/mips/float_dsp_mips.c | 41 ++-
> 7 files changed, 1248 insertions(+), 16 deletions(-)
> create mode 100644 libavcodec/mips/aacdec_mips.c
> create mode 100644 libavcodec/mips/aacdec_mips.h
>
> diff --git a/libavcodec/aac.h b/libavcodec/aac.h
> index 9e9fa0e..7b96e3b 100644
> --- a/libavcodec/aac.h
> +++ b/libavcodec/aac.h
> @@ -257,10 +257,12 @@ typedef struct ChannelElement {
> SpectralBandReplication sbr;
> } ChannelElement;
>
> +typedef struct AACContext AACContext;
> +
> /**
> * main AAC context
> */
> -typedef struct AACContext {
> +struct AACContext {
> AVClass *class;
> AVCodecContext *avctx;
> AVFrame frame;
> @@ -318,6 +320,18 @@ typedef struct AACContext {
>
> OutputConfiguration oc[2];
> int warned_num_aac_frames;
> -} AACContext;
> +
> + /* aacdec functions pointers */
> + void (*imdct_and_windowing)(AACContext *ac, SingleChannelElement *sce);
> + void (*apply_ltp)(AACContext *ac, SingleChannelElement *sce);
> + void (*apply_tns)(float coef[1024], TemporalNoiseShaping *tns,
> + IndividualChannelStream *ics, int decode);
> + void (*windowing_and_mdct_ltp)(AACContext *ac, float *out,
> + float *in, IndividualChannelStream *ics);
> + void (*update_ltp)(AACContext *ac, SingleChannelElement *sce);
> +
> +};
> +
> +extern VLC vlc_spectral[11];
>
> #endif /* AVCODEC_AAC_H */
> diff --git a/libavcodec/aacdec.c b/libavcodec/aacdec.c
> index 239153a..4beacc9 100644
> --- a/libavcodec/aacdec.c
> +++ b/libavcodec/aacdec.c
> @@ -108,10 +108,12 @@
>
> #if ARCH_ARM
> # include "arm/aac.h"
> +#elif ARCH_MIPS
> +# include "mips/aacdec_mips.h"
> #endif
>
> static VLC vlc_scalefactors;
> -static VLC vlc_spectral[11];
> +VLC vlc_spectral[11];
global variables need a ff_ prefix
but can you explain what needs this table to be global ?
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Many things microsoft did are stupid, but not doing something just because
microsoft did it is even more stupid. If everything ms did were stupid they
would be bankrupt already.
More information about the ffmpeg-devel
mailing list