add corresponding .S files for BaseMemoryLibXXX X64 arch. Note that we use ".intel_syntax noprefix" directive to simplify the .S implementation.

git-svn-id: https://edk2.svn.sourceforge.net/svnroot/edk2/trunk/edk2@6184 6f19259b-4bc3-4df7-8a09-765794883524

MdePkg/Library/BaseMemoryLibSse2/X64/ZeroMem.S

@@ -0,0 +1,66 @@
+#
+# ConvertAsm.py: Automatically generated from ZeroMem.asm
+#
+#------------------------------------------------------------------------------
+#
+# Copyright (c) 2006, Intel Corporation
+# All rights reserved. This program and the accompanying materials
+# are licensed and made available under the terms and conditions of the BSD License
+# which accompanies this distribution.  The full text of the license may be found at
+# http://opensource.org/licenses/bsd-license.php
+#
+# THE PROGRAM IS DISTRIBUTED UNDER THE BSD LICENSE ON AN "AS IS" BASIS,
+# WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED.
+#
+# Module Name:
+#
+#   ZeroMem.S
+#
+# Abstract:
+#
+#   ZeroMem function
+#
+# Notes:
+#
+#------------------------------------------------------------------------------
+
+
+#------------------------------------------------------------------------------
+#  VOID *
+#  InternalMemZeroMem (
+#    IN VOID   *Buffer,
+#    IN UINTN  Count
+#    );
+#------------------------------------------------------------------------------
+.intel_syntax noprefix
+.globl ASM_PFX(InternalMemZeroMem)
+ASM_PFX(InternalMemZeroMem):
+    push    rdi
+    mov     rdi, rcx
+    xor     rcx, rcx
+    xor     eax, eax
+    sub     rcx, rdi
+    and     rcx, 15
+    mov     r8, rdi
+    jz      L0
+    cmp     rcx, rdx
+    cmova   rcx, rdx
+    sub     rdx, rcx
+    rep     stosb
+L0:
+    mov     rcx, rdx
+    and     edx, 15
+    shr     rcx, 4
+    jz      _ZeroBytes_L2
+    pxor    xmm0, xmm0
+L1:
+    movntdq [rdi], xmm0                 # rdi should be 16-byte aligned
+    add     rdi, 16
+    loop    L1
+    mfence
+_ZeroBytes_L2:
+    mov     ecx, edx
+    rep     stosb
+    mov     rax, r8
+    pop     rdi
+    ret