diff options
| author | Martin Storsjö <martin@martin.st> | 2019-10-07 13:29:41 +0300 |
|---|---|---|
| committer | Jean-Baptiste Kempf <jb@videolan.org> | 2019-10-08 09:18:25 +0300 |
| commit | 61442bee60f45b05da627ddbac10a9a63e243f47 (patch) | |
| tree | fc7bb14ebeb763b87e063c80d7f7897b21a8dba1 /src/arm/32/util.S | |
| parent | 5647a57eabc454e2e2360429aba494452af00cb3 (diff) | |
arm: mc: Port the ARM64 warp filter to arm32
Relative speedup over C code:
Cortex A7 A8 A9 A53 A72 A73
warp_8x8_8bpc_neon: 2.79 5.45 4.18 3.96 4.16 4.51
warp_8x8t_8bpc_neon: 2.79 5.33 4.18 3.98 4.22 4.25
Comparison to original ARM64 assembly:
ARM64: Cortex A53 A72 A73
warp_8x8_8bpc_neon: 1854.6 1072.5 1102.5
warp_8x8t_8bpc_neon: 1839.6 1069.4 1089.5
ARM32:
warp_8x8_8bpc_neon: 2132.5 1160.3 1218.0
warp_8x8t_8bpc_neon: 2113.7 1148.0 1209.1
Diffstat (limited to 'src/arm/32/util.S')
| -rw-r--r-- | src/arm/32/util.S | 15 |
1 files changed, 15 insertions, 0 deletions
diff --git a/src/arm/32/util.S b/src/arm/32/util.S index 3f2e328..ddf157f 100644 --- a/src/arm/32/util.S +++ b/src/arm/32/util.S @@ -69,4 +69,19 @@ #endif .endm +.macro transpose_8x8b q0, q1, q2, q3, r0, r1, r2, r3, r4, r5, r6, r7 + vtrn.32 \q0, \q2 + vtrn.32 \q1, \q3 + + vtrn.16 \r0, \r2 + vtrn.16 \r1, \r3 + vtrn.16 \r4, \r6 + vtrn.16 \r5, \r7 + + vtrn.8 \r0, \r1 + vtrn.8 \r2, \r3 + vtrn.8 \r4, \r5 + vtrn.8 \r6, \r7 +.endm + #endif /* DAV1D_SRC_ARM_32_UTIL_S */ |