TextureDecoder: Pass the TLUT address straight into the texture decoder

This removes the requirement for the TextureDecoder to have access to
global texture memory.

6 files changed, 76 insertions, 74 deletions

diff --git a/Source/Core/VideoBackends/Software/TextureSampler.cpp b/Source/Core/VideoBackends/Software/TextureSampler.cpp
index 9629a7600b..cc90f11ae2 100644
--- a/Source/Core/VideoBackends/Software/TextureSampler.cpp
+++ b/Source/Core/VideoBackends/Software/TextureSampler.cpp
@@ -106,6 +106,7 @@ void SampleMip(s32 s, s32 t, s32 mip, bool linear, u8 texmap, u8 *sample)
 	TexMode0& tm0 = texUnit.texMode0[subTexmap];
 	TexImage0& ti0 = texUnit.texImage0[subTexmap];
 	TexTLUT& texTlut = texUnit.texTlut[subTexmap];
+	TlutFormat tlutfmt = (TlutFormat) texTlut.tlut_format;
 
 	u8 *imageSrc, *imageSrcOdd = nullptr;
 	if (texUnit.texImage1[subTexmap].image_type)
@@ -124,6 +125,7 @@ void SampleMip(s32 s, s32 t, s32 mip, bool linear, u8 texmap, u8 *sample)
 	int imageHeight = ti0.height;
 
 	int tlutAddress = texTlut.tmem_offset << 9;
+	const u8* tlut = &texMem[tlutAddress];
 
 	// reduce sample location and texture size to mip level
 	// move texture pointer to mip location
@@ -179,20 +181,18 @@ void SampleMip(s32 s, s32 t, s32 mip, bool linear, u8 texmap, u8 *sample)
 		WrapCoord(imageSPlus1, tm0.wrap_s, imageWidth);
 		WrapCoord(imageTPlus1, tm0.wrap_t, imageHeight);
 
-		TlutFormat tlutfmt = (TlutFormat) texTlut.tlut_format;
-
 		if (!(ti0.format == GX_TF_RGBA8 && texUnit.texImage1[subTexmap].image_type))
 		{
-			TexDecoder_DecodeTexel(sampledTex, imageSrc, imageS, imageT, imageWidth, ti0.format, tlutAddress, tlutfmt);
+			TexDecoder_DecodeTexel(sampledTex, imageSrc, imageS, imageT, imageWidth, ti0.format, tlut, tlutfmt);
 			SetTexel(sampledTex, texel, (128 - fractS) * (128 - fractT));
 
-			TexDecoder_DecodeTexel(sampledTex, imageSrc, imageSPlus1, imageT, imageWidth, ti0.format, tlutAddress, tlutfmt);
+			TexDecoder_DecodeTexel(sampledTex, imageSrc, imageSPlus1, imageT, imageWidth, ti0.format, tlut, tlutfmt);
 			AddTexel(sampledTex, texel, (fractS) * (128 - fractT));
 
-			TexDecoder_DecodeTexel(sampledTex, imageSrc, imageS, imageTPlus1, imageWidth, ti0.format, tlutAddress, tlutfmt);
+			TexDecoder_DecodeTexel(sampledTex, imageSrc, imageS, imageTPlus1, imageWidth, ti0.format, tlut, tlutfmt);
 			AddTexel(sampledTex, texel, (128 - fractS) * (fractT));
 
-			TexDecoder_DecodeTexel(sampledTex, imageSrc, imageSPlus1, imageTPlus1, imageWidth, ti0.format, tlutAddress, tlutfmt);
+			TexDecoder_DecodeTexel(sampledTex, imageSrc, imageSPlus1, imageTPlus1, imageWidth, ti0.format, tlut, tlutfmt);
 			AddTexel(sampledTex, texel, (fractS) * (fractT));
 		}
 		else
@@ -226,7 +226,7 @@ void SampleMip(s32 s, s32 t, s32 mip, bool linear, u8 texmap, u8 *sample)
 		WrapCoord(imageT, tm0.wrap_t, imageHeight);
 
 		if (!(ti0.format == GX_TF_RGBA8 && texUnit.texImage1[subTexmap].image_type))
-			TexDecoder_DecodeTexel(sample, imageSrc, imageS, imageT, imageWidth, ti0.format, tlutAddress, (TlutFormat) texTlut.tlut_format);
+			TexDecoder_DecodeTexel(sample, imageSrc, imageS, imageT, imageWidth, ti0.format, tlut, tlutfmt);
 		else
 			TexDecoder_DecodeTexelRGBA8FromTmem(sample, imageSrc, imageSrcOdd, imageS, imageT, imageWidth);
 	}
diff --git a/Source/Core/VideoCommon/TextureCacheBase.cpp b/Source/Core/VideoCommon/TextureCacheBase.cpp
index cbfdb47d62..c0b46d92b6 100644
--- a/Source/Core/VideoCommon/TextureCacheBase.cpp
+++ b/Source/Core/VideoCommon/TextureCacheBase.cpp
@@ -490,7 +490,8 @@ TextureCache::TCacheEntryBase* TextureCache::Load(unsigned int const stage,
 	{
 		if (!(texformat == GX_TF_RGBA8 && from_tmem))
 		{
-			pcfmt = TexDecoder_Decode(temp, src_data, expandedWidth, expandedHeight, texformat, tlutaddr, (TlutFormat) tlutfmt);
+			const u8* tlut = &texMem[tlutaddr];
+			pcfmt = TexDecoder_Decode(temp, src_data, expandedWidth, expandedHeight, texformat, tlut, (TlutFormat) tlutfmt);
 		}
 		else
 		{
@@ -566,7 +567,8 @@ TextureCache::TCacheEntryBase* TextureCache::Load(unsigned int const stage,
 				const u8*& mip_src_data = from_tmem
 					? ((level % 2) ? ptr_odd : ptr_even)
 					: src_data;
-				TexDecoder_Decode(temp, mip_src_data, expanded_mip_width, expanded_mip_height, texformat, tlutaddr, (TlutFormat) tlutfmt);
+				const u8* tlut = &texMem[tlutaddr];
+				TexDecoder_Decode(temp, mip_src_data, expanded_mip_width, expanded_mip_height, texformat, tlut, (TlutFormat) tlutfmt);
 				mip_src_data += TexDecoder_GetTextureSizeInBytes(expanded_mip_width, expanded_mip_height, texformat);
 
 				entry->Load(mip_width, mip_height, expanded_mip_width, level);
diff --git a/Source/Core/VideoCommon/TextureDecoder.h b/Source/Core/VideoCommon/TextureDecoder.h
index 9838881ce6..be566fcf16 100644
--- a/Source/Core/VideoCommon/TextureDecoder.h
+++ b/Source/Core/VideoCommon/TextureDecoder.h
@@ -78,12 +78,12 @@ enum PC_TexFormat
 	PC_TEX_FMT_DXT1,
 };
 
-PC_TexFormat TexDecoder_Decode(u8 *dst, const u8 *src, int width, int height, int texformat, int tlutaddr, TlutFormat tlutfmt);
+PC_TexFormat TexDecoder_Decode(u8 *dst, const u8 *src, int width, int height, int texformat, const u8* tlut, TlutFormat tlutfmt);
 PC_TexFormat TexDecoder_DecodeRGBA8FromTmem(u8* dst, const u8 *src_ar, const u8 *src_gb, int width, int height);
-void TexDecoder_DecodeTexel(u8 *dst, const u8 *src, int s, int t, int imageWidth, int texformat, int tlutaddr, TlutFormat tlutfmt);
+void TexDecoder_DecodeTexel(u8 *dst, const u8 *src, int s, int t, int imageWidth, int texformat, const u8* tlut, TlutFormat tlutfmt);
 void TexDecoder_DecodeTexelRGBA8FromTmem(u8 *dst, const u8 *src_ar, const u8* src_gb, int s, int t, int imageWidth);
 
 void TexDecoder_SetTexFmtOverlayOptions(bool enable, bool center);
 
 /* Internal method, implemented by TextureDecoder_Generic and TextureDecoder_x64. */
-PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int height, int texformat, int tlutaddr, TlutFormat tlutfmt);
+PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int height, int texformat, const u8* tlut, TlutFormat tlutfmt);
diff --git a/Source/Core/VideoCommon/TextureDecoder_Common.cpp b/Source/Core/VideoCommon/TextureDecoder_Common.cpp
index 4cb8167104..cb71c3613f 100644
--- a/Source/Core/VideoCommon/TextureDecoder_Common.cpp
+++ b/Source/Core/VideoCommon/TextureDecoder_Common.cpp
@@ -242,9 +242,9 @@ static void TexDecoder_DrawOverlay(u8 *dst, int width, int height, int texformat
 	}
 }
 
-PC_TexFormat TexDecoder_Decode(u8 *dst, const u8 *src, int width, int height, int texformat, int tlutaddr, TlutFormat tlutfmt)
+PC_TexFormat TexDecoder_Decode(u8 *dst, const u8 *src, int width, int height, int texformat, const u8* tlut, TlutFormat tlutfmt)
 {
-	PC_TexFormat pc_texformat = _TexDecoder_DecodeImpl((u32*)dst, src, width, height, texformat, tlutaddr, tlutfmt);
+	PC_TexFormat pc_texformat = _TexDecoder_DecodeImpl((u32*)dst, src, width, height, texformat, tlut, tlutfmt);
 
 	if (TexFmt_Overlay_Enable && pc_texformat != PC_TEX_FMT_NONE)
 		TexDecoder_DrawOverlay(dst, width, height, texformat, pc_texformat);
@@ -301,7 +301,7 @@ struct DXTBlock
 	u8 lines[4];
 };
 
-void TexDecoder_DecodeTexel(u8 *dst, const u8 *src, int s, int t, int imageWidth, int texformat, int tlutaddr, TlutFormat tlutfmt)
+void TexDecoder_DecodeTexel(u8 *dst, const u8 *src, int s, int t, int imageWidth, int texformat, const u8* tlut_, TlutFormat tlutfmt)
 {
 	/* General formula for computing texture offset
 	//
@@ -330,7 +330,7 @@ void TexDecoder_DecodeTexel(u8 *dst, const u8 *src, int s, int t, int imageWidth
 			u32 offset = base + (blkOff >> 1);
 
 			u8 val = (*(src + offset) >> rs) & 0xF;
-			u16 *tlut = (u16*)(texMem + tlutaddr);
+			u16 *tlut = (u16*) tlut_;
 
 			switch (tlutfmt)
 			{
@@ -395,7 +395,7 @@ void TexDecoder_DecodeTexel(u8 *dst, const u8 *src, int s, int t, int imageWidth
 			u32 blkOff = (blkT << 3) + blkS;
 
 			u8 val = *(src + base + blkOff);
-			u16 *tlut = (u16*)(texMem + tlutaddr);
+			u16 *tlut = (u16*) tlut_;
 
 			switch (tlutfmt)
 			{
@@ -460,7 +460,7 @@ void TexDecoder_DecodeTexel(u8 *dst, const u8 *src, int s, int t, int imageWidth
 			const u16* valAddr = (u16*)(src + offset);
 
 			u16 val = Common::swap16(*valAddr) & 0x3FFF;
-			u16 *tlut = (u16*)(texMem + tlutaddr);
+			u16 *tlut = (u16*) tlut_;
 
 			switch (tlutfmt)
 			{
diff --git a/Source/Core/VideoCommon/TextureDecoder_Generic.cpp b/Source/Core/VideoCommon/TextureDecoder_Generic.cpp
index 3dff970afa..41144b79b5 100644
--- a/Source/Core/VideoCommon/TextureDecoder_Generic.cpp
+++ b/Source/Core/VideoCommon/TextureDecoder_Generic.cpp
@@ -60,9 +60,9 @@ struct DXTBlock
 	u8 lines[4];
 };
 
-static inline void decodebytesC4_5A3_To_rgba32(u32 *dst, const u8 *src, int tlutaddr)
+static inline void decodebytesC4_5A3_To_rgba32(u32 *dst, const u8 *src, const u8* tlut_)
 {
-	u16 *tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 4; x++)
 	{
 		u8 val = src[x];
@@ -71,9 +71,9 @@ static inline void decodebytesC4_5A3_To_rgba32(u32 *dst, const u8 *src, int tlut
 	}
 }
 
-static inline void decodebytesC4IA8_To_RGBA(u32* dst, const u8* src, int tlutaddr)
+static inline void decodebytesC4IA8_To_RGBA(u32* dst, const u8* src, const u8* tlut_)
 {
-	u16* tlut = (u16*)(texMem+tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 4; x++)
 	{
 		u8 val = src[x];
@@ -82,9 +82,9 @@ static inline void decodebytesC4IA8_To_RGBA(u32* dst, const u8* src, int tlutadd
 	}
 }
 
-static inline void decodebytesC4RGB565_To_RGBA(u32* dst, const u8* src, int tlutaddr)
+static inline void decodebytesC4RGB565_To_RGBA(u32* dst, const u8* src, const u8* tlut_)
 {
-	u16* tlut = (u16*)(texMem+tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 4; x++)
 	{
 		u8 val = src[x];
@@ -93,9 +93,9 @@ static inline void decodebytesC4RGB565_To_RGBA(u32* dst, const u8* src, int tlut
 	}
 }
 
-static inline void decodebytesC8_5A3_To_RGBA32(u32 *dst, const u8 *src, int tlutaddr)
+static inline void decodebytesC8_5A3_To_RGBA32(u32 *dst, const u8 *src, const u8* tlut_)
 {
-	u16 *tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 8; x++)
 	{
 		u8 val = src[x];
@@ -103,27 +103,27 @@ static inline void decodebytesC8_5A3_To_RGBA32(u32 *dst, const u8 *src, int tlut
 	}
 }
 
-static inline void decodebytesC8IA8_To_RGBA(u32* dst, const u8* src, int tlutaddr)
+static inline void decodebytesC8IA8_To_RGBA(u32* dst, const u8* src, const u8* tlut_)
 {
-	u16* tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 8; x++)
 	{
 		*dst++ = decodeIA8Swapped(tlut[src[x]]);
 	}
 }
 
-static inline void decodebytesC8RGB565_To_RGBA(u32* dst, const u8* src, int tlutaddr)
+static inline void decodebytesC8RGB565_To_RGBA(u32* dst, const u8* src, const u8* tlut_)
 {
-	u16* tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 8; x++)
 	{
 		*dst++ = decode565RGBA(Common::swap16(tlut[src[x]]));
 	}
 }
 
-static inline void decodebytesC14X2_5A3_To_RGBA(u32 *dst, const u16 *src, int tlutaddr)
+static inline void decodebytesC14X2_5A3_To_RGBA(u32 *dst, const u16 *src, const u8* tlut_)
 {
-	u16 *tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 4; x++)
 	{
 		u16 val = Common::swap16(src[x]);
@@ -131,9 +131,9 @@ static inline void decodebytesC14X2_5A3_To_RGBA(u32 *dst, const u16 *src, int tl
 	}
 }
 
-static inline void decodebytesC14X2IA8_To_RGBA(u32* dst, const u16* src, int tlutaddr)
+static inline void decodebytesC14X2IA8_To_RGBA(u32* dst, const u16* src, const u8* tlut_)
 {
-	u16* tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 4; x++)
 	{
 		u16 val = Common::swap16(src[x]);
@@ -141,9 +141,9 @@ static inline void decodebytesC14X2IA8_To_RGBA(u32* dst, const u16* src, int tlu
 	}
 }
 
-static inline void decodebytesC14X2rgb565_To_RGBA(u32* dst, const u16* src, int tlutaddr)
+static inline void decodebytesC14X2rgb565_To_RGBA(u32* dst, const u16* src, const u8* tlut_)
 {
-	u16* tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 4; x++)
 	{
 		u16 val = Common::swap16(src[x]);
@@ -246,7 +246,7 @@ static void decodeDXTBlockRGBA(u32 *dst, const DXTBlock *src, int pitch)
 // TODO: complete SSE2 optimization of less often used texture formats.
 // TODO: refactor algorithms using _mm_loadl_epi64 unaligned loads to prefer 128-bit aligned loads.
 
-PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int height, int texformat, int tlutaddr, TlutFormat tlutfmt)
+PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int height, int texformat, const u8* tlut, TlutFormat tlutfmt)
 {
 
 	const int Wsteps4 = (width + 3) / 4;
@@ -261,21 +261,21 @@ PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int he
 			for (int y = 0; y < height; y += 8)
 				for (int x = 0, yStep = (y / 8) * Wsteps8; x < width; x += 8,yStep++)
 					for (int iy = 0, xStep =  8 * yStep; iy < 8; iy++,xStep++)
-						decodebytesC4_5A3_To_rgba32(dst + (y + iy) * width + x, src + 4 * xStep, tlutaddr);
+						decodebytesC4_5A3_To_rgba32(dst + (y + iy) * width + x, src + 4 * xStep, tlut);
 		}
 		else if (tlutfmt == GX_TL_IA8)
 		{
 			for (int y = 0; y < height; y += 8)
 				for (int x = 0, yStep = (y / 8) * Wsteps8; x < width; x += 8,yStep++)
 					for (int iy = 0, xStep =  8 * yStep; iy < 8; iy++,xStep++)
-						decodebytesC4IA8_To_RGBA(dst + (y + iy) * width + x, src + 4 * xStep, tlutaddr);
+						decodebytesC4IA8_To_RGBA(dst + (y + iy) * width + x, src + 4 * xStep, tlut);
 		}
 		else if (tlutfmt == GX_TL_RGB565)
 		{
 			for (int y = 0; y < height; y += 8)
 				for (int x = 0, yStep = (y / 8) * Wsteps8; x < width; x += 8,yStep++)
 					for (int iy = 0, xStep =  8 * yStep; iy < 8; iy++,xStep++)
-						decodebytesC4RGB565_To_RGBA(dst + (y + iy) * width + x, src  + 4 * xStep, tlutaddr);
+						decodebytesC4RGB565_To_RGBA(dst + (y + iy) * width + x, src  + 4 * xStep, tlut);
 		}
 		break;
 	case GX_TF_I4:
@@ -323,21 +323,21 @@ PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int he
 			for (int y = 0; y < height; y += 4)
 				for (int x = 0, yStep = (y / 4) * Wsteps8; x < width; x += 8, yStep++)
 					for (int iy = 0, xStep = 4 * yStep; iy < 4; iy++, xStep++)
-						decodebytesC8_5A3_To_RGBA32((u32*)dst + (y + iy) * width + x, src + 8 * xStep, tlutaddr);
+						decodebytesC8_5A3_To_RGBA32((u32*)dst + (y + iy) * width + x, src + 8 * xStep, tlut);
 		}
 		else if (tlutfmt == GX_TL_IA8)
 		{
 			for (int y = 0; y < height; y += 4)
 				for (int x = 0, yStep = (y / 4) * Wsteps8; x < width; x += 8, yStep++)
 					for (int iy = 0, xStep = 4 * yStep; iy < 4; iy++, xStep++)
-						decodebytesC8IA8_To_RGBA(dst + (y + iy) * width + x, src + 8 * xStep, tlutaddr);
+						decodebytesC8IA8_To_RGBA(dst + (y + iy) * width + x, src + 8 * xStep, tlut);
 		}
 		else if (tlutfmt == GX_TL_RGB565)
 		{
 			for (int y = 0; y < height; y += 4)
 				for (int x = 0, yStep = (y / 4) * Wsteps8; x < width; x += 8, yStep++)
 					for (int iy = 0, xStep = 4 * yStep; iy < 4; iy++, xStep++)
-						decodebytesC8RGB565_To_RGBA(dst + (y + iy) * width + x, src + 8 * xStep, tlutaddr);
+						decodebytesC8RGB565_To_RGBA(dst + (y + iy) * width + x, src + 8 * xStep, tlut);
 
 		}
 		break;
@@ -372,21 +372,21 @@ PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int he
 			for (int y = 0; y < height; y += 4)
 				for (int x = 0, yStep = (y / 4) * Wsteps4; x < width; x += 4, yStep++)
 					for (int iy = 0, xStep = 4 * yStep; iy < 4; iy++, xStep++)
-						decodebytesC14X2_5A3_To_RGBA(dst + (y + iy) * width + x, (u16*)(src + 8 * xStep), tlutaddr);
+						decodebytesC14X2_5A3_To_RGBA(dst + (y + iy) * width + x, (u16*)(src + 8 * xStep), tlut);
 		}
 		else if (tlutfmt == GX_TL_IA8)
 		{
 			for (int y = 0; y < height; y += 4)
 				for (int x = 0, yStep = (y / 4) * Wsteps4; x < width; x += 4, yStep++)
 					for (int iy = 0, xStep = 4 * yStep; iy < 4; iy++, xStep++)
-						decodebytesC14X2IA8_To_RGBA(dst + (y + iy) * width + x,  (u16*)(src + 8 * xStep), tlutaddr);
+						decodebytesC14X2IA8_To_RGBA(dst + (y + iy) * width + x,  (u16*)(src + 8 * xStep), tlut);
 		}
 		else if (tlutfmt == GX_TL_RGB565)
 		{
 			for (int y = 0; y < height; y += 4)
 				for (int x = 0, yStep = (y / 4) * Wsteps4; x < width; x += 4, yStep++)
 					for (int iy = 0, xStep = 4 * yStep; iy < 4; iy++, xStep++)
-						decodebytesC14X2rgb565_To_RGBA(dst + (y + iy) * width + x, (u16*)(src + 8 * xStep), tlutaddr);
+						decodebytesC14X2rgb565_To_RGBA(dst + (y + iy) * width + x, (u16*)(src + 8 * xStep), tlut);
 		}
 		break;
 	case GX_TF_RGB565:
diff --git a/Source/Core/VideoCommon/TextureDecoder_x64.cpp b/Source/Core/VideoCommon/TextureDecoder_x64.cpp
index 867566032b..9603ec1695 100644
--- a/Source/Core/VideoCommon/TextureDecoder_x64.cpp
+++ b/Source/Core/VideoCommon/TextureDecoder_x64.cpp
@@ -83,9 +83,9 @@ struct DXTBlock
 	u8 lines[4];
 };
 
-static inline void decodebytesC4_5A3_To_rgba32(u32 *dst, const u8 *src, int tlutaddr)
+static inline void decodebytesC4_5A3_To_rgba32(u32 *dst, const u8 *src, const u8* tlut_)
 {
-	u16 *tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 4; x++)
 	{
 		u8 val = src[x];
@@ -94,9 +94,9 @@ static inline void decodebytesC4_5A3_To_rgba32(u32 *dst, const u8 *src, int tlut
 	}
 }
 
-static inline void decodebytesC4IA8_To_RGBA(u32* dst, const u8* src, int tlutaddr)
+static inline void decodebytesC4IA8_To_RGBA(u32* dst, const u8* src, const u8* tlut_)
 {
-	u16* tlut = (u16*)(texMem+tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 4; x++)
 	{
 		u8 val = src[x];
@@ -105,9 +105,9 @@ static inline void decodebytesC4IA8_To_RGBA(u32* dst, const u8* src, int tlutadd
 	}
 }
 
-static inline void decodebytesC4RGB565_To_RGBA(u32* dst, const u8* src, int tlutaddr)
+static inline void decodebytesC4RGB565_To_RGBA(u32* dst, const u8* src, const u8* tlut_)
 {
-	u16* tlut = (u16*)(texMem+tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 4; x++)
 	{
 		u8 val = src[x];
@@ -116,9 +116,9 @@ static inline void decodebytesC4RGB565_To_RGBA(u32* dst, const u8* src, int tlut
 	}
 }
 
-static inline void decodebytesC8_5A3_To_RGBA32(u32 *dst, const u8 *src, int tlutaddr)
+static inline void decodebytesC8_5A3_To_RGBA32(u32 *dst, const u8 *src, const u8* tlut_)
 {
-	u16 *tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 8; x++)
 	{
 		u8 val = src[x];
@@ -126,27 +126,27 @@ static inline void decodebytesC8_5A3_To_RGBA32(u32 *dst, const u8 *src, int tlut
 	}
 }
 
-static inline void decodebytesC8IA8_To_RGBA(u32* dst, const u8* src, int tlutaddr)
+static inline void decodebytesC8IA8_To_RGBA(u32* dst, const u8* src, const u8* tlut_)
 {
-	u16* tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 8; x++)
 	{
 		*dst++ = decodeIA8Swapped(tlut[src[x]]);
 	}
 }
 
-static inline void decodebytesC8RGB565_To_RGBA(u32* dst, const u8* src, int tlutaddr)
+static inline void decodebytesC8RGB565_To_RGBA(u32* dst, const u8* src, const u8* tlut_)
 {
-	u16* tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 8; x++)
 	{
 		*dst++ = decode565RGBA(Common::swap16(tlut[src[x]]));
 	}
 }
 
-static inline void decodebytesC14X2_5A3_To_RGBA(u32 *dst, const u16 *src, int tlutaddr)
+static inline void decodebytesC14X2_5A3_To_RGBA(u32 *dst, const u16 *src, const u8* tlut_)
 {
-	u16 *tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 4; x++)
 	{
 		u16 val = Common::swap16(src[x]);
@@ -154,9 +154,9 @@ static inline void decodebytesC14X2_5A3_To_RGBA(u32 *dst, const u16 *src, int tl
 	}
 }
 
-static inline void decodebytesC14X2IA8_To_RGBA(u32* dst, const u16* src, int tlutaddr)
+static inline void decodebytesC14X2IA8_To_RGBA(u32* dst, const u16* src, const u8* tlut_)
 {
-	u16* tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 4; x++)
 	{
 		u16 val = Common::swap16(src[x]);
@@ -164,9 +164,9 @@ static inline void decodebytesC14X2IA8_To_RGBA(u32* dst, const u16* src, int tlu
 	}
 }
 
-static inline void decodebytesC14X2rgb565_To_RGBA(u32* dst, const u16* src, int tlutaddr)
+static inline void decodebytesC14X2rgb565_To_RGBA(u32* dst, const u16* src, const u8* tlut_)
 {
-	u16* tlut = (u16*)(texMem + tlutaddr);
+	const u16* tlut = (u16*) tlut_;
 	for (int x = 0; x < 4; x++)
 	{
 		u16 val = Common::swap16(src[x]);
@@ -287,7 +287,7 @@ static inline void SetOpenMPThreadCount(int width, int height)
 // TODO: complete SSE2 optimization of less often used texture formats.
 // TODO: refactor algorithms using _mm_loadl_epi64 unaligned loads to prefer 128-bit aligned loads.
 
-PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int height, int texformat, int tlutaddr, TlutFormat tlutfmt)
+PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int height, int texformat, const u8* tlut, TlutFormat tlutfmt)
 {
 	SetOpenMPThreadCount(width, height);
 
@@ -304,7 +304,7 @@ PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int he
 			for (int y = 0; y < height; y += 8)
 				for (int x = 0, yStep = (y / 8) * Wsteps8; x < width; x += 8,yStep++)
 					for (int iy = 0, xStep =  8 * yStep; iy < 8; iy++,xStep++)
-						decodebytesC4_5A3_To_rgba32(dst + (y + iy) * width + x, src + 4 * xStep, tlutaddr);
+						decodebytesC4_5A3_To_rgba32(dst + (y + iy) * width + x, src + 4 * xStep, tlut);
 		}
 		else if (tlutfmt == GX_TL_IA8)
 		{
@@ -312,7 +312,7 @@ PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int he
 			for (int y = 0; y < height; y += 8)
 				for (int x = 0, yStep = (y / 8) * Wsteps8; x < width; x += 8,yStep++)
 					for (int iy = 0, xStep =  8 * yStep; iy < 8; iy++,xStep++)
-						decodebytesC4IA8_To_RGBA(dst + (y + iy) * width + x, src + 4 * xStep, tlutaddr);
+						decodebytesC4IA8_To_RGBA(dst + (y + iy) * width + x, src + 4 * xStep, tlut);
 
 		}
 		else if (tlutfmt == GX_TL_RGB565)
@@ -321,7 +321,7 @@ PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int he
 			for (int y = 0; y < height; y += 8)
 				for (int x = 0, yStep = (y / 8) * Wsteps8; x < width; x += 8,yStep++)
 					for (int iy = 0, xStep =  8 * yStep; iy < 8; iy++,xStep++)
-						decodebytesC4RGB565_To_RGBA(dst + (y + iy) * width + x, src  + 4 * xStep, tlutaddr);
+						decodebytesC4RGB565_To_RGBA(dst + (y + iy) * width + x, src  + 4 * xStep, tlut);
 		}
 		break;
 	case GX_TF_I4:
@@ -562,7 +562,7 @@ PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int he
 			for (int y = 0; y < height; y += 4)
 				for (int x = 0, yStep = (y / 4) * Wsteps8; x < width; x += 8, yStep++)
 					for (int iy = 0, xStep = 4 * yStep; iy < 4; iy++, xStep++)
-						decodebytesC8_5A3_To_RGBA32((u32*)dst + (y + iy) * width + x, src + 8 * xStep, tlutaddr);
+						decodebytesC8_5A3_To_RGBA32((u32*)dst + (y + iy) * width + x, src + 8 * xStep, tlut);
 		}
 		else if (tlutfmt == GX_TL_IA8)
 		{
@@ -570,7 +570,7 @@ PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int he
 			for (int y = 0; y < height; y += 4)
 					for (int x = 0, yStep = (y / 4) * Wsteps8; x < width; x += 8, yStep++)
 						for (int iy = 0, xStep = 4 * yStep; iy < 4; iy++, xStep++)
-							decodebytesC8IA8_To_RGBA(dst + (y + iy) * width + x, src + 8 * xStep, tlutaddr);
+							decodebytesC8IA8_To_RGBA(dst + (y + iy) * width + x, src + 8 * xStep, tlut);
 
 		}
 		else if (tlutfmt == GX_TL_RGB565)
@@ -579,7 +579,7 @@ PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int he
 			for (int y = 0; y < height; y += 4)
 					for (int x = 0, yStep = (y / 4) * Wsteps8; x < width; x += 8, yStep++)
 						for (int iy = 0, xStep = 4 * yStep; iy < 4; iy++, xStep++)
-							decodebytesC8RGB565_To_RGBA(dst + (y + iy) * width + x, src + 8 * xStep, tlutaddr);
+							decodebytesC8RGB565_To_RGBA(dst + (y + iy) * width + x, src + 8 * xStep, tlut);
 
 		}
 		break;
@@ -675,7 +675,7 @@ PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int he
 			for (int y = 0; y < height; y += 4)
 				for (int x = 0, yStep = (y / 4) * Wsteps4; x < width; x += 4, yStep++)
 					for (int iy = 0, xStep = 4 * yStep; iy < 4; iy++, xStep++)
-						decodebytesC14X2_5A3_To_RGBA(dst + (y + iy) * width + x, (u16*)(src + 8 * xStep), tlutaddr);
+						decodebytesC14X2_5A3_To_RGBA(dst + (y + iy) * width + x, (u16*)(src + 8 * xStep), tlut);
 		}
 		else if (tlutfmt == GX_TL_IA8)
 		{
@@ -683,7 +683,7 @@ PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int he
 			for (int y = 0; y < height; y += 4)
 				for (int x = 0, yStep = (y / 4) * Wsteps4; x < width; x += 4, yStep++)
 					for (int iy = 0, xStep = 4 * yStep; iy < 4; iy++, xStep++)
-						decodebytesC14X2IA8_To_RGBA(dst + (y + iy) * width + x,  (u16*)(src + 8 * xStep), tlutaddr);
+						decodebytesC14X2IA8_To_RGBA(dst + (y + iy) * width + x,  (u16*)(src + 8 * xStep), tlut);
 		}
 		else if (tlutfmt == GX_TL_RGB565)
 		{
@@ -691,7 +691,7 @@ PC_TexFormat _TexDecoder_DecodeImpl(u32 * dst, const u8 * src, int width, int he
 			for (int y = 0; y < height; y += 4)
 				for (int x = 0, yStep = (y / 4) * Wsteps4; x < width; x += 4, yStep++)
 					for (int iy = 0, xStep = 4 * yStep; iy < 4; iy++, xStep++)
-						decodebytesC14X2rgb565_To_RGBA(dst + (y + iy) * width + x, (u16*)(src + 8 * xStep), tlutaddr);
+						decodebytesC14X2rgb565_To_RGBA(dst + (y + iy) * width + x, (u16*)(src + 8 * xStep), tlut);
 		}
 		break;
 	case GX_TF_RGB565: