diff --git a/src/Cafe/GamePatch.cpp b/src/Cafe/GamePatch.cpp
index 3f90e94a..f197d011 100644
--- a/src/Cafe/GamePatch.cpp
+++ b/src/Cafe/GamePatch.cpp
@@ -245,6 +245,10 @@ uint8 botw_crashFuncMask[] = { 0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF
 
 uint8 ffl_floatArrayEndianSwap[] = { 0x7C,0x08,0x02,0xA6,0x94,0x21,0xFF,0xE8,0x93,0xC1,0x00,0x10,0x7C,0x7E,0x1B,0x78,0x93,0xE1,0x00,0x14,0x93,0x81,0x00,0x08,0x7C,0x9F,0x23,0x78,0x93,0xA1,0x00,0x0C,0x90,0x01,0x00,0x1C,0x3B,0xA0,0x00,0x00,0x7C,0x1D,0xF8,0x40,0x40,0x80,0x00,0x20,0x57,0xBC,0x10,0x3A,0x7C,0x3E,0xE4,0x2E };
 
+// Alternative FFL endian swap signature (Mario Kart 8 and possibly other titles)
+// Same algorithm as above but with a different compiler-scheduled prologue ordering
+uint8 ffl_floatArrayEndianSwap_v2[] = { 0x7C,0x08,0x02,0xA6,0x94,0x21,0xFF,0xE8,0x93,0xE1,0x00,0x14,0x93,0xA1,0x00,0x0C,0x7C,0x9F,0x23,0x78,0x93,0x81,0x00,0x08,0x93,0xC1,0x00,0x10,0x90,0x01,0x00,0x1C,0x3B,0xA0,0x00,0x00,0x7C,0x1D,0xF8,0x40,0x7C,0x7E,0x1B,0x78,0x40,0x80,0x00,0x20,0x57,0xBC,0x10,0x3A,0x7C,0x3E,0xE4,0x2E };
+
 uint8 xcx_enterCriticalSectionSignature[] = { 0x94,0x21,0xFF,0xE0,0xBF,0x41,0x00,0x08,0x7C,0x08,0x02,0xA6,0x90,0x01,0x00,0x24,0x7C,0x7E,0x1B,0x78,0x80,0x1E,0x00,0x08,0x2C,0x00,0x00,0x00,0x41,0x82,0x00,0xC0,0x48,0x01,0xD7,0xA1,0x7C,0x7A,0x1B,0x79,0x41,0x82,0x00,0xB4,0x81,0x3E,0x00,0x04,0x7C,0x09,0xD0,0x40,0x40,0x82,0x00,0x2C,0x7D,0x20,0xF0,0x28,0x7C,0x00,0xF0,0x6C };
 uint8 xcx_enterCriticalSectionMask[] =      { 0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0x00,0x00,0x00,0x00,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF,0xFF };
 
@@ -361,11 +365,23 @@ void GamePatch_scan()
 	hleAddr = hle_locate(ffl_floatArrayEndianSwap, NULL, sizeof(ffl_floatArrayEndianSwap));
 	if (hleAddr)
 	{
-		cemuLog_logDebug(LogType::Force, "HLE: Hook FFL float array endian swap function at 0x{:08x}", hleAddr);
+		cemuLog_logDebug(LogType::Force, "HLE: Hook FFL float array endian swap function (v1) at 0x{:08x}", hleAddr);
 		sint32 functionIndex = hleIndex_h000000003;
 		uint32 opcode = (1 << 26) | (functionIndex); // opcode for HLE: 0x1000 + FunctionIndex
 		memory_write<uint32>(hleAddr, opcode);
 	}
+	else
+	{
+		// Try v2 signature (Mario Kart 8 and possibly other titles with reordered prologue)
+		hleAddr = hle_locate(ffl_floatArrayEndianSwap_v2, NULL, sizeof(ffl_floatArrayEndianSwap_v2));
+		if (hleAddr)
+		{
+			cemuLog_logDebug(LogType::Force, "HLE: Hook FFL float array endian swap function (v2) at 0x{:08x}", hleAddr);
+			sint32 functionIndex = hleIndex_h000000003;
+			uint32 opcode = (1 << 26) | (functionIndex); // opcode for HLE: 0x1000 + FunctionIndex
+			memory_write<uint32>(hleAddr, opcode);
+		}
+	}
 
 	// XCX freeze workaround
 	//hleAddr = hle_locate(xcx_enterCriticalSectionSignature, xcx_enterCriticalSectionMask, sizeof(xcx_enterCriticalSectionSignature));