--- /dev/null
+;------------------------------------------------------------------------------\r
+;\r
+; Copyright (c) 2006, Intel Corporation\r
+; All rights reserved. This program and the accompanying materials\r
+; are licensed and made available under the terms and conditions of the BSD License\r
+; which accompanies this distribution. The full text of the license may be found at\r
+; http://opensource.org/licenses/bsd-license.php\r
+;\r
+; THE PROGRAM IS DISTRIBUTED UNDER THE BSD LICENSE ON AN "AS IS" BASIS,\r
+; WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED.\r
+;\r
+; Module Name:\r
+;\r
+; CopyMem.asm\r
+;\r
+; Abstract:\r
+;\r
+; CopyMem function\r
+;\r
+; Notes:\r
+;\r
+;------------------------------------------------------------------------------\r
+\r
+ .686\r
+ .model flat,C\r
+ .xmm\r
+ .code\r
+\r
+;------------------------------------------------------------------------------\r
+; VOID *\r
+; InternalMemCopyMem (\r
+; IN VOID *Destination,\r
+; IN VOID *Source,\r
+; IN UINTN Count\r
+; );\r
+;------------------------------------------------------------------------------\r
+InternalMemCopyMem PROC USES esi edi\r
+ mov esi, [esp + 16] ; esi <- Source\r
+ mov edi, [esp + 12] ; edi <- Destination\r
+ mov edx, [esp + 20] ; edx <- Count\r
+ lea eax, [esi + edx - 1] ; eax <- End of Source\r
+ cmp esi, edi\r
+ jae @F\r
+ cmp eax, edi ; Overlapped?\r
+ jae @CopyBackward ; Copy backward if overlapped\r
+@@:\r
+ xor ecx, ecx\r
+ sub ecx, edi\r
+ and ecx, 15 ; ecx + edi aligns on 16-byte boundary\r
+ jz @F\r
+ cmp ecx, edx\r
+ cmova ecx, edx\r
+ sub edx, ecx ; edx <- remaining bytes to copy\r
+ rep movsb\r
+@@:\r
+ mov ecx, edx\r
+ and edx, 15\r
+ shr ecx, 4 ; ecx <- # of DQwords to copy\r
+ jz @CopyBytes\r
+ add esp, -16\r
+ movdqu [esp], xmm0 ; save xmm0\r
+@@:\r
+ movdqu xmm0, [esi] ; esi may not be 16-bytes aligned\r
+ movntdq [edi], xmm0 ; edi should be 16-bytes aligned\r
+ add esi, 16\r
+ add edi, 16\r
+ loop @B\r
+ mfence\r
+ movdqu xmm0, [esp] ; restore xmm0\r
+ add esp, 16 ; stack cleanup\r
+ jmp @CopyBytes\r
+@CopyBackward:\r
+ mov esi, eax ; esi <- Last byte in Source\r
+ lea edi, [edi + edx - 1] ; edi <- Last byte in Destination\r
+ std\r
+@CopyBytes:\r
+ mov ecx, edx\r
+ rep movsb\r
+ cld\r
+ mov eax, [esp + 12] ; eax <- Destination as return value\r
+ ret\r
+InternalMemCopyMem ENDP\r
+\r
+ END\r