On Sat, 11 Jan 2014, Janne Grunau wrote:
Ported from ARMv7 NEON.
---
libavcodec/aarch64/Makefile | 3 +-
libavcodec/aarch64/h264dsp_init_aarch64.c | 14 ++
libavcodec/aarch64/h264dsp_neon.S | 261 ++++++++++++++++++++++++++++++
libavcodec/aarch64/neon.S | 46 ++++--
4 files changed, 312 insertions(+), 12 deletions(-)
create mode 100644 libavcodec/aarch64/h264dsp_neon.S
diff --git a/libavcodec/aarch64/neon.S b/libavcodec/aarch64/neon.S
index 2449cfe..f1072b7 100644
--- a/libavcodec/aarch64/neon.S
+++ b/libavcodec/aarch64/neon.S
@@ -48,9 +48,9 @@
trn2 \r7\().2S, \r9\().2S, \r7\().2S
.endm
-.macro transpose_8x8B2 r0, r1, r2, r3, r4, r5, r6, r7, r8, r9
- trn1 \r8\().16B, \r0\().16B, \r1\().16B
- trn2 \r9\().16B, \r0\().16B, \r1\().16B
+.macro transpose_8x16B r0, r1, r2, r3, r4, r5, r6, r7, t0, t1
+ trn1 \t0\().16B, \r0\().16B, \r1\().16B
+ trn2 \t1\().16B, \r0\().16B, \r1\().16B
trn1 \r1\().16B, \r2\().16B, \r3\().16B
trn2 \r3\().16B, \r2\().16B, \r3\().16B
trn1 \r0\().16B, \r4\().16B, \r5\().16B
Why this sudden change of macro arguments here?
The rest of it looks ok although I didn't read the translated asm line by
line.
// Martin
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel
