From c5a786ed2c27ed0669f9518ce9e7a6e499a09d0c Mon Sep 17 00:00:00 2001
From: Gregory Hainaut <gregory.hainaut@gmail.com>
Date: Fri, 14 Aug 2015 20:57:45 +0200
Subject: [PATCH] gsdx-ogl: remove support WMS/T == 2 in hardware unit

I think behavior was wrong because only first texel coordinate was clamped.

Beside we can't interpolate if AEM isn't yet applied
---
 plugins/GSdx/GSRendererOGL.cpp    |  2 +-
 plugins/GSdx/res/glsl/tfx_fs.glsl | 38 ++++++++++++++-----------------
 plugins/GSdx/res/glsl_source.h    | 38 ++++++++++++++-----------------
 3 files changed, 35 insertions(+), 43 deletions(-)

diff --git a/plugins/GSdx/GSRendererOGL.cpp b/plugins/GSdx/GSRendererOGL.cpp
index 043684b00c..1b9e122954 100644
--- a/plugins/GSdx/GSRendererOGL.cpp
+++ b/plugins/GSdx/GSRendererOGL.cpp
@@ -781,7 +781,7 @@ void GSRendererOGL::DrawPrims(GSTexture* rt, GSTexture* ds, GSTextureCache::Sour
 		const GSLocalMemory::psm_t &psm = GSLocalMemory::m_psm[m_context->TEX0.PSM];
 		const GSLocalMemory::psm_t &cpsm = psm.pal > 0 ? GSLocalMemory::m_psm[m_context->TEX0.CPSM] : psm;
 		bool bilinear = m_filter == 2 ? m_vt.IsLinear() : m_filter != 0;
-		bool simple_sample = !tex->m_palette && cpsm.fmt == 0 && m_context->CLAMP.WMS < 3 && m_context->CLAMP.WMT < 3;
+		bool simple_sample = !tex->m_palette && cpsm.fmt == 0 && m_context->CLAMP.WMS < 2 && m_context->CLAMP.WMT < 2;
 		// Don't force extra filtering on sprite (it creates various upscaling issue)
 		bilinear &= !((m_vt.m_primclass == GS_SPRITE_CLASS) && m_userhacks_round_sprite_offset && !m_vt.IsLinear());
 
diff --git a/plugins/GSdx/res/glsl/tfx_fs.glsl b/plugins/GSdx/res/glsl/tfx_fs.glsl
index df482558ed..beac76b0f1 100644
--- a/plugins/GSdx/res/glsl/tfx_fs.glsl
+++ b/plugins/GSdx/res/glsl/tfx_fs.glsl
@@ -63,16 +63,25 @@ layout(std140, binding = 21) uniform cb21
 {
 	vec3 FogColor;
 	float AREF;
+
 	vec4 WH;
+
 	vec2 MinF;
 	vec2 TA;
+
 	uvec4 MskFix;
+
 	uvec4 FbMask;
-	vec3 _not_yet_used;
+
+	vec3 _pad1;
 	float Af;
+
 	vec4 HalfTexel;
+
 	vec4 MinMax;
+
 	vec2 TC_OffsetHack;
+	vec2 _pad2;
 };
 
 vec4 sample_c(vec2 uv)
@@ -85,7 +94,7 @@ vec4 sample_p(float idx)
 	return texture(PaletteSampler, vec2(idx, 0.0f));
 }
 
-vec4 wrapuv(vec4 uv)
+vec4 clamp_wrap_uv(vec4 uv)
 {
 	vec4 uv_out = uv;
 
@@ -120,21 +129,6 @@ vec4 wrapuv(vec4 uv)
 	return uv_out;
 }
 
-vec2 clampuv(vec2 uv)
-{
-	vec2 uv_out = uv;
-
-#if (PS_WMS == 2) && (PS_WMT == 2)
-	uv_out = clamp(uv, MinF, MinMax.zw);
-#elif PS_WMS == 2
-	uv_out.x = clamp(uv.x, MinF.x, MinMax.z);
-#elif PS_WMT == 2
-	uv_out.y = clamp(uv.y, MinF.y, MinMax.w);
-#endif
-
-	return uv_out;
-}
-
 mat4 sample_4c(vec4 uv)
 {
 	mat4 c;
@@ -211,10 +205,12 @@ vec4 sample_color(vec2 st, float q)
 	mat4 c;
 	vec2 dd;
 
-#if (PS_LTF == 0 && PS_FMT <= FMT_16 && PS_WMS < 3 && PS_WMT < 3)
-	c[0] = sample_c(clampuv(st));
+    // FIXME I'm not sure this condition is useful (I think code will be optimized)
+#if (PS_LTF == 0 && PS_FMT == FMT_32 && PS_WMS < 2 && PS_WMT < 2)
+	// No software LTF and pure 32 bits RGBA texure without special texture wrapping
+	c[0] = sample_c(st);
 #ifdef TEX_COORD_DEBUG
-	c[0].rg = clampuv(st).xy;
+	c[0].rg = st.xy;
 #endif
 
 #else
@@ -230,7 +226,7 @@ vec4 sample_color(vec2 st, float q)
 		uv = st.xyxy;
 	}
 
-	uv = wrapuv(uv);
+	uv = clamp_wrap_uv(uv);
 
 	if((PS_FMT & FMT_PAL) != 0)
 	{
diff --git a/plugins/GSdx/res/glsl_source.h b/plugins/GSdx/res/glsl_source.h
index 3139c86853..413c7bd8fd 100644
--- a/plugins/GSdx/res/glsl_source.h
+++ b/plugins/GSdx/res/glsl_source.h
@@ -922,16 +922,25 @@ static const char* tfx_fs_all_glsl =
 	"{\n"
 	"	vec3 FogColor;\n"
 	"	float AREF;\n"
+	"\n"
 	"	vec4 WH;\n"
+	"\n"
 	"	vec2 MinF;\n"
 	"	vec2 TA;\n"
+	"\n"
 	"	uvec4 MskFix;\n"
+	"\n"
 	"	uvec4 FbMask;\n"
-	"	vec3 _not_yet_used;\n"
+	"\n"
+	"	vec3 _pad1;\n"
 	"	float Af;\n"
+	"\n"
 	"	vec4 HalfTexel;\n"
+	"\n"
 	"	vec4 MinMax;\n"
+	"\n"
 	"	vec2 TC_OffsetHack;\n"
+	"	vec2 _pad2;\n"
 	"};\n"
 	"\n"
 	"vec4 sample_c(vec2 uv)\n"
@@ -944,7 +953,7 @@ static const char* tfx_fs_all_glsl =
 	"	return texture(PaletteSampler, vec2(idx, 0.0f));\n"
 	"}\n"
 	"\n"
-	"vec4 wrapuv(vec4 uv)\n"
+	"vec4 clamp_wrap_uv(vec4 uv)\n"
 	"{\n"
 	"	vec4 uv_out = uv;\n"
 	"\n"
@@ -979,21 +988,6 @@ static const char* tfx_fs_all_glsl =
 	"	return uv_out;\n"
 	"}\n"
 	"\n"
-	"vec2 clampuv(vec2 uv)\n"
-	"{\n"
-	"	vec2 uv_out = uv;\n"
-	"\n"
-	"#if (PS_WMS == 2) && (PS_WMT == 2)\n"
-	"	uv_out = clamp(uv, MinF, MinMax.zw);\n"
-	"#elif PS_WMS == 2\n"
-	"	uv_out.x = clamp(uv.x, MinF.x, MinMax.z);\n"
-	"#elif PS_WMT == 2\n"
-	"	uv_out.y = clamp(uv.y, MinF.y, MinMax.w);\n"
-	"#endif\n"
-	"\n"
-	"	return uv_out;\n"
-	"}\n"
-	"\n"
 	"mat4 sample_4c(vec4 uv)\n"
 	"{\n"
 	"	mat4 c;\n"
@@ -1070,10 +1064,12 @@ static const char* tfx_fs_all_glsl =
 	"	mat4 c;\n"
 	"	vec2 dd;\n"
 	"\n"
-	"#if (PS_LTF == 0 && PS_FMT <= FMT_16 && PS_WMS < 3 && PS_WMT < 3)\n"
-	"	c[0] = sample_c(clampuv(st));\n"
+	"    // FIXME I'm not sure this condition is useful (I think code will be optimized)\n"
+	"#if (PS_LTF == 0 && PS_FMT == FMT_32 && PS_WMS < 2 && PS_WMT < 2)\n"
+	"	// No software LTF and pure 32 bits RGBA texure without special texture wrapping\n"
+	"	c[0] = sample_c(st);\n"
 	"#ifdef TEX_COORD_DEBUG\n"
-	"	c[0].rg = clampuv(st).xy;\n"
+	"	c[0].rg = st.xy;\n"
 	"#endif\n"
 	"\n"
 	"#else\n"
@@ -1089,7 +1085,7 @@ static const char* tfx_fs_all_glsl =
 	"		uv = st.xyxy;\n"
 	"	}\n"
 	"\n"
-	"	uv = wrapuv(uv);\n"
+	"	uv = clamp_wrap_uv(uv);\n"
 	"\n"
 	"	if((PS_FMT & FMT_PAL) != 0)\n"
 	"	{\n"