Author Topic: YMM copy.  (Read 1161 times)


  • Administrator
  • Member
  • ******
  • Posts: 6105
  • Mnemonic Driven API Grinder
    • The MASM32 SDK
YMM copy.
« on: August 22, 2016, 11:39:40 AM »
I thought someone may like this. Seems to be a bit faster than an XMM version in at least some contexts. I just tweaked the second instruction to "vmovntdq" and it is now clearly faster than the XMM version.

; ¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤


ymmcopya proc

  ; rcx = source address
  ; rdx = destination address
  ; r8  = byte count

    mov r11, r8
    shr r11, 5                  ; div by 32 for loop count
    xor r10, r10                ; zero r10 to use as index

    vmovdqa ymm0, YMMWORD PTR [rcx+r10]
    vmovntdq YMMWORD PTR [rdx+r10], ymm0

    add r10, 32
    sub r11, 1
    jnz lpst

    mov rax, r8                 ; calculate remainder if any
    and rax, 31
    test rax, rax
    jnz @F

    mov r9b, [rcx+r10]          ; copy any remainder
    mov [rdx+r10], r9b
    add r10, 1
    sub rax, 1
    jnz @B


ymmcopya endp


; ¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤
hutch at movsd dot com    :biggrin:  :biggrin: