+ volatile UINT32 *Destination32;\r
+ CONST UINT32 *Source32;\r
+ volatile UINT64 *Destination64;\r
+ CONST UINT64 *Source64;\r
+ UINTN Alignment;\r
+\r
+ if ((((UINTN)DestinationBuffer & 0x7) == 0) && (((UINTN)SourceBuffer & 0x7) == 0) && (Length >= 8)) {\r
+ if (SourceBuffer > DestinationBuffer) {\r
+ Destination64 = (UINT64*)DestinationBuffer;\r
+ Source64 = (CONST UINT64*)SourceBuffer;\r
+ while (Length >= 8) {\r
+ *(Destination64++) = *(Source64++);\r
+ Length -= 8;\r
+ }\r
+\r
+ // Finish if there are still some bytes to copy\r
+ Destination8 = (UINT8*)Destination64;\r
+ Source8 = (CONST UINT8*)Source64;\r
+ while (Length-- != 0) {\r
+ *(Destination8++) = *(Source8++);\r
+ }\r
+ } else if (SourceBuffer < DestinationBuffer) {\r
+ Destination64 = (UINT64*)((UINTN)DestinationBuffer + Length);\r
+ Source64 = (CONST UINT64*)((UINTN)SourceBuffer + Length);\r
+\r
+ // Destination64 and Source64 were aligned on a 64-bit boundary\r
+ // but if length is not a multiple of 8 bytes then they won't be\r
+ // anymore.\r
+\r
+ Alignment = Length & 0x7;\r
+ if (Alignment != 0) {\r
+ Destination8 = (UINT8*)Destination64;\r
+ Source8 = (CONST UINT8*)Source64;\r
+\r
+ while (Alignment-- != 0) {\r
+ *(--Destination8) = *(--Source8);\r
+ --Length;\r
+ }\r
+ Destination64 = (UINT64*)Destination8;\r
+ Source64 = (CONST UINT64*)Source8;\r
+ }\r
+\r
+ while (Length > 0) {\r
+ *(--Destination64) = *(--Source64);\r
+ Length -= 8;\r
+ }\r
+ }\r
+ } else if ((((UINTN)DestinationBuffer & 0x3) == 0) && (((UINTN)SourceBuffer & 0x3) == 0) && (Length >= 4)) {\r
+ if (SourceBuffer > DestinationBuffer) {\r
+ Destination32 = (UINT32*)DestinationBuffer;\r
+ Source32 = (CONST UINT32*)SourceBuffer;\r
+ while (Length >= 4) {\r
+ *(Destination32++) = *(Source32++);\r
+ Length -= 4;\r
+ }\r
+\r
+ // Finish if there are still some bytes to copy\r
+ Destination8 = (UINT8*)Destination32;\r
+ Source8 = (CONST UINT8*)Source32;\r
+ while (Length-- != 0) {\r
+ *(Destination8++) = *(Source8++);\r
+ }\r
+ } else if (SourceBuffer < DestinationBuffer) {\r
+ Destination32 = (UINT32*)((UINTN)DestinationBuffer + Length);\r
+ Source32 = (CONST UINT32*)((UINTN)SourceBuffer + Length);\r
+\r
+ // Destination32 and Source32 were aligned on a 32-bit boundary\r
+ // but if length is not a multiple of 4 bytes then they won't be\r
+ // anymore.\r
+\r
+ Alignment = Length & 0x3;\r
+ if (Alignment != 0) {\r
+ Destination8 = (UINT8*)Destination32;\r
+ Source8 = (CONST UINT8*)Source32;\r
+\r
+ while (Alignment-- != 0) {\r
+ *(--Destination8) = *(--Source8);\r
+ --Length;\r
+ }\r
+ Destination32 = (UINT32*)Destination8;\r
+ Source32 = (CONST UINT32*)Source8;\r
+ }\r