From 3f835f563fbacf424331b2f164c88bde7bc77adc Mon Sep 17 00:00:00 2001
From: 1e1 <1e1@users.noreply.github.com>
Date: Thu, 2 Jul 2026 22:40:59 +0200
Subject: [PATCH] Espresso: use std::countr_zero for register-set bit scans

Replace the byte-then-bit linear scan in IMLPhysRegisterSet::GetFirstAvailableReg / GetNextAvailableReg with std::countr_zero (single tzcnt / rbit+clz). Behaviour-identical (both return the index of the lowest set bit), branchless, and ~16x faster on the isolated scan in a micro-benchmark - it removes the mispredicted inner-loop branches.
---
 .../Recompiler/IML/IMLRegisterAllocator.h     | 30 ++-----------------
 1 file changed, 3 insertions(+), 27 deletions(-)

diff --git a/src/Cafe/HW/Espresso/Recompiler/IML/IMLRegisterAllocator.h b/src/Cafe/HW/Espresso/Recompiler/IML/IMLRegisterAllocator.h
index 0a54e4cb..97899616 100644
--- a/src/Cafe/HW/Espresso/Recompiler/IML/IMLRegisterAllocator.h
+++ b/src/Cafe/HW/Espresso/Recompiler/IML/IMLRegisterAllocator.h
@@ -63,19 +63,7 @@ public:
 	IMLPhysReg GetFirstAvailableReg()
 	{
 		cemu_assert_debug(m_regBitmask != 0);
-		sint32 regIndex = 0;
-		auto tmp = m_regBitmask;
-		while ((tmp & 0xFF) == 0)
-		{
-			regIndex += 8;
-			tmp >>= 8;
-		}
-		while ((tmp & 0x1) == 0)
-		{
-			regIndex++;
-			tmp >>= 1;
-		}
-		return regIndex;
+		return (IMLPhysReg)std::countr_zero(m_regBitmask);
 	}
 
 	// returns index of next available register (search includes any register index >= startIndex)
@@ -84,22 +72,10 @@ public:
 	{
 		if (startIndex >= 64)
 			return -1;
-		uint32 regIndex = startIndex;
-		auto tmp = m_regBitmask;
-		tmp >>= regIndex;
+		uint64 tmp = m_regBitmask >> startIndex;
 		if (!tmp)
 			return -1;
-		while ((tmp & 0xFF) == 0)
-		{
-			regIndex += 8;
-			tmp >>= 8;
-		}
-		while ((tmp & 0x1) == 0)
-		{
-			regIndex++;
-			tmp >>= 1;
-		}
-		return regIndex;
+		return startIndex + (IMLPhysReg)std::countr_zero(tmp);
 	}
 
 	sint32 CountAvailableRegs() const