diff options
| author | Lynne <dev@lynne.ee> | 2021-02-27 04:11:04 +0100 |
|---|---|---|
| committer | Lynne <dev@lynne.ee> | 2021-02-27 04:21:05 +0100 |
| commit | 8e94b7cff03539bcb4c360d2550a031a5378df03 (patch) | |
| tree | d22c567906393c503244f1d43caa4d9c500528fe /libavutil/tx_priv.h | |
| parent | 9ddaf0c9f06fab9194161425a32615c4cfc2ec20 (diff) | |
| download | ffmpeg-8e94b7cff03539bcb4c360d2550a031a5378df03.tar.gz | |
lavu/tx: invert permutation lookups
out[lut[i]] = in[i] lookups were 4.04 times(!) slower than
out[i] = in[lut[i]] lookups for an out-of-place FFT of length 4096.
The permutes remain unchanged for anything but out-of-place monolithic
FFT, as those benefit quite a lot from the current order (it means
there's only 1 lookup necessary to add to an offset, rather than
a full gather).
The code was based around non-power-of-two FFTs, so this wasn't
benchmarked early on.
Diffstat (limited to 'libavutil/tx_priv.h')
| -rw-r--r-- | libavutil/tx_priv.h | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/libavutil/tx_priv.h b/libavutil/tx_priv.h index e9fba02a35..e2f4314a4f 100644 --- a/libavutil/tx_priv.h +++ b/libavutil/tx_priv.h @@ -123,7 +123,7 @@ struct AVTXContext { /* Shared functions */ int ff_tx_type_is_mdct(enum AVTXType type); int ff_tx_gen_compound_mapping(AVTXContext *s); -int ff_tx_gen_ptwo_revtab(AVTXContext *s); +int ff_tx_gen_ptwo_revtab(AVTXContext *s, int invert_lookup); int ff_tx_gen_ptwo_inplace_revtab_idx(AVTXContext *s); /* Also used by SIMD init */ |