FD.io VPP
v16.12-rc0-308-g931be3a
Vector Packet Processing
|
Go to the source code of this file.
Typedefs | |
typedef u8 | u8x16u |
typedef u8 | u8x32u |
Functions | |
static void | clib_mov16 (u8 *dst, const u8 *src) |
static void | clib_mov32 (u8 *dst, const u8 *src) |
static void | clib_mov64 (u8 *dst, const u8 *src) |
static void | clib_mov128 (u8 *dst, const u8 *src) |
static void | clib_mov256 (u8 *dst, const u8 *src) |
static void | clib_mov64blocks (u8 *dst, const u8 *src, size_t n) |
static void | clib_mov256blocks (u8 *dst, const u8 *src, size_t n) |
static void * | clib_memcpy (void *dst, const void *src, size_t n) |
Definition at line 54 of file memcpy_avx.h.
Definition at line 55 of file memcpy_avx.h.
|
inlinestatic |
Copy less than 16 bytes
Fast way when copy size doesn't exceed 512 bytes
Make store aligned when copy size exceeds 512 bytes
Copy 256-byte blocks. Use copy block function for better instruction order control, which is important when load is unaligned.
Copy 64-byte blocks. Use copy block function for better instruction order control, which is important when load is unaligned.
Copy whatever left
Definition at line 147 of file memcpy_avx.h.
Definition at line 77 of file memcpy_avx.h.
Definition at line 84 of file memcpy_avx.h.
Definition at line 70 of file memcpy_avx.h.