ligang/vpp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Replacing vtbl NEON intrinsic with rev NEON intrinsic for byte_swap.

Browse Source 
Using rev16 vector intrinsic to reverse byteorder in each word
independently.

Change-Id: I071c40780baffe0bda614ec5d9dd92858f574b0d
Signed-off-by: Sirshak Das <sirshak.das@arm.com>
Reviewed-by: Steve Capper <steve.capper@arm.com>
Reviewed-by: Brian Brooks <brian.brooks@arm.com>
Reviewed-by: Yi He <yi.he@arm.com>
Verified-by: Lijian Zhang <lijian.zhang@arm.com>
This commit is contained in:
Sirshak Das
2018-08-22 12:02:04 +08:00
committed by Damjan Marion
parent 8e5d5db24e
commit 6da42decd4
1 changed files with 1 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
src/vppinfra/vector_neon.h
View File

@@ -121,11 +121,7 @@ foreach_neon_vec128i foreach_neon_vec128u
static_always_inline u16x8
u16x8_byte_swap (u16x8 v)
{
const u8 swap_pattern[] = {
1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14,
};
u8x16 swap = vld1q_u8 (swap_pattern);
return (u16x8) vqtbl1q_u8 ((u8x16) v, swap);
return (u16x8) vrev16q_u8 ((u8x16) v);
}
static_always_inline u8x16