[FFmpeg-cvslog] x86/tx_float: enable SIMD for sizes over 131072

Lynne git at videolan.org
Wed Feb 7 16:21:29 EET 2024


ffmpeg | branch: master | Lynne <dev at lynne.ee> | Thu Jan 18 17:30:29 2024 +0100| [bd3e71b21ec3786ec6fc00ef260af0150f31b71b] | committer: Lynne

x86/tx_float: enable SIMD for sizes over 131072

The tables for the new sizes were added last year due
to being required for SDR.
However, the assembly was never updated to use them.

> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=bd3e71b21ec3786ec6fc00ef260af0150f31b71b
---

 libavutil/x86/tx_float.asm | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/libavutil/x86/tx_float.asm b/libavutil/x86/tx_float.asm
index e1533a8595..42006848f1 100644
--- a/libavutil/x86/tx_float.asm
+++ b/libavutil/x86/tx_float.asm
@@ -46,7 +46,7 @@
 %endif
 
 %assign i 16
-%rep 14
+%rep 18
 cextern tab_ %+ i %+ _float ; ff_tab_i_float...
 %assign i (i << 1)
 %endrep
@@ -1385,7 +1385,11 @@ FFT_SPLIT_RADIX_DEF 8192,  .16384pt
 FFT_SPLIT_RADIX_DEF 16384, .32768pt
 FFT_SPLIT_RADIX_DEF 32768, .65536pt
 FFT_SPLIT_RADIX_DEF 65536, .131072pt
-FFT_SPLIT_RADIX_DEF 131072
+FFT_SPLIT_RADIX_DEF 131072, .262144pt
+FFT_SPLIT_RADIX_DEF 262144, .524288pt
+FFT_SPLIT_RADIX_DEF 524288, .1048576pt
+FFT_SPLIT_RADIX_DEF 1048576, .2097152pt
+FFT_SPLIT_RADIX_DEF 2097152
 
 ;===============================================================================
 ; Final synthesis + deinterleaving code



More information about the ffmpeg-cvslog mailing list