Add group_alpha plane to draw device and painters.

3 files changed, 877 insertions, 543 deletions

diff --git a/source/fitz/draw-affine.c b/source/fitz/draw-affine.c
index db6c2cd5..c64411b4 100644
--- a/source/fitz/draw-affine.c
+++ b/source/fitz/draw-affine.c
@@ -7,7 +7,7 @@
 
 typedef unsigned char byte;
 
-typedef void (paintfn_t)(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop);
+typedef void (paintfn_t)(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop);
 
 static inline int lerp(int a, int b, int t)
 {
@@ -31,7 +31,7 @@ static inline const byte *sample_nearest(const byte *s, int w, int h, int str, i
 /* Blend premultiplied source image in constant alpha over destination */
 
 static inline void
-template_affine_alpha_N_lerp(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, byte * restrict hp)
+template_affine_alpha_N_lerp(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, byte * restrict hp, byte * restrict gp)
 {
 	int k;
 
@@ -47,7 +47,8 @@ template_affine_alpha_N_lerp(byte * restrict dp, int da, const byte * restrict s
 			const byte *b = sample_nearest(sp, sw, sh, ss, sn1+sa, ui+1, vi);
 			const byte *c = sample_nearest(sp, sw, sh, ss, sn1+sa, ui, vi+1);
 			const byte *d = sample_nearest(sp, sw, sh, ss, sn1+sa, ui+1, vi+1);
-			int xa = sa ? fz_mul255(bilerp(a[sn1], b[sn1], c[sn1], d[sn1], uf, vf), alpha) : alpha;
+			int x = sa ? bilerp(a[sn1], b[sn1], c[sn1], d[sn1], uf, vf) : 255;
+			int xa = sa ? fz_mul255(x, alpha) : alpha;
 			if (xa != 0)
 			{
 				int t = 255 - xa;
@@ -61,12 +62,16 @@ template_affine_alpha_N_lerp(byte * restrict dp, int da, const byte * restrict s
 				if (da)
 					dp[dn1] = xa + fz_mul255(dp[dn1], t);
 				if (hp)
-					hp[0] = xa + fz_mul255(hp[0], t);
+					hp[0] = x + fz_mul255(hp[0], 255 - x);
+				if (gp)
+					gp[0] = xa + fz_mul255(gp[0], t);
 			}
 		}
 		dp += dn1+da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -74,7 +79,7 @@ template_affine_alpha_N_lerp(byte * restrict dp, int da, const byte * restrict s
 }
 
 static inline void
-template_affine_alpha_N_lerp_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, byte * restrict hp, const fz_overprint * restrict eop)
+template_affine_alpha_N_lerp_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	int k;
 
@@ -90,7 +95,8 @@ template_affine_alpha_N_lerp_op(byte * restrict dp, int da, const byte * restric
 			const byte *b = sample_nearest(sp, sw, sh, ss, sn1+sa, ui+1, vi);
 			const byte *c = sample_nearest(sp, sw, sh, ss, sn1+sa, ui, vi+1);
 			const byte *d = sample_nearest(sp, sw, sh, ss, sn1+sa, ui+1, vi+1);
-			int xa = sa ? fz_mul255(bilerp(a[sn1], b[sn1], c[sn1], d[sn1], uf, vf), alpha) : alpha;
+			int x = sa ? bilerp(a[sn1], b[sn1], c[sn1], d[sn1], uf, vf) : 255;
+			int xa = sa ? fz_mul255(x, alpha) : alpha;
 			if (xa != 0)
 			{
 				int t = 255 - xa;
@@ -108,12 +114,16 @@ template_affine_alpha_N_lerp_op(byte * restrict dp, int da, const byte * restric
 				if (da)
 					dp[dn1] = xa + fz_mul255(dp[dn1], t);
 				if (hp)
-					hp[0] = xa + fz_mul255(hp[0], t);
+					hp[0] = x + fz_mul255(hp[0], 255-x);
+				if (gp)
+					gp[0] = xa + fz_mul255(gp[0], t);
 			}
 		}
 		dp += dn1+da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -122,7 +132,7 @@ template_affine_alpha_N_lerp_op(byte * restrict dp, int da, const byte * restric
 
 /* Special case code for gray -> rgb */
 static inline void
-template_affine_alpha_g2rgb_lerp(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int alpha, byte * restrict hp)
+template_affine_alpha_g2rgb_lerp(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int alpha, byte * restrict hp, byte * restrict gp)
 {
 	do
 	{
@@ -136,24 +146,29 @@ template_affine_alpha_g2rgb_lerp(byte * restrict dp, int da, const byte * restri
 			const byte *b = sample_nearest(sp, sw, sh, ss, 1+sa, ui+1, vi);
 			const byte *c = sample_nearest(sp, sw, sh, ss, 1+sa, ui, vi+1);
 			const byte *d = sample_nearest(sp, sw, sh, ss, 1+sa, ui+1, vi+1);
-			int y = (sa ? fz_mul255(bilerp(a[1], b[1], c[1], d[1], uf, vf), alpha) : alpha);
-			if (y != 0)
+			int y = sa ? bilerp(a[1], b[1], c[1], d[1], uf, vf) : 255;
+			int ya = (sa ? fz_mul255(y, alpha) : alpha);
+			if (ya != 0)
 			{
 				int x = bilerp(a[0], b[0], c[0], d[0], uf, vf);
-				int t = 255 - y;
+				int t = 255 - ya;
 				x = fz_mul255(x, alpha);
 				dp[0] = x + fz_mul255(dp[0], t);
 				dp[1] = x + fz_mul255(dp[1], t);
 				dp[2] = x + fz_mul255(dp[2], t);
 				if (da)
-					dp[3] = y + fz_mul255(dp[3], t);
+					dp[3] = ya + fz_mul255(dp[3], t);
 				if (hp)
-					hp[0] = y + fz_mul255(hp[0], t);
+					hp[0] = y + fz_mul255(hp[0], 255 - y);
+				if (gp)
+					gp[0] = ya + fz_mul255(gp[0], t);
 			}
 		}
 		dp += 4;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -161,7 +176,7 @@ template_affine_alpha_g2rgb_lerp(byte * restrict dp, int da, const byte * restri
 }
 
 static inline void
-template_affine_alpha_N_near_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, byte * restrict hp)
+template_affine_alpha_N_near_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, byte * restrict hp, byte * restrict gp)
 {
 	int k;
 	int ui = u >> 16;
@@ -175,30 +190,35 @@ template_affine_alpha_N_near_fa0(byte * restrict dp, int da, const byte * restri
 		if (vi >= 0 && vi < sh)
 		{
 			const byte *sample = sp + (vi * ss);
-			int a = (sa ? fz_mul255(sample[sn1], alpha) : alpha);
-			if (a != 0)
+			int a = sa ? sample[sn1] : 255;
+			int aa = (sa ? fz_mul255(a, alpha) : alpha);
+			if (aa != 0)
 			{
-				int t = 255 - a;
+				int t = 255 - aa;
 				for (k = 0; k < sn1; k++)
 					dp[k] = fz_mul255(sample[k], alpha) + fz_mul255(dp[k], t);
 				for (; k < dn1; k++)
 					dp[k] = 0;
 				if (da)
-					dp[dn1] = a + fz_mul255(dp[dn1], t);
+					dp[dn1] = aa + fz_mul255(dp[dn1], t);
 				if (hp)
-					hp[0] = a + fz_mul255(hp[0], t);
+					hp[0] = a + fz_mul255(hp[0], 255 - a);
+				if (gp)
+					gp[0] = aa + fz_mul255(gp[0], t);
 			}
 		}
 		dp += dn1+da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		v += fb;
 	}
 	while (--w);
 }
 
 static inline void
-template_affine_alpha_N_near_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, byte * restrict hp)
+template_affine_alpha_N_near_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, byte * restrict hp, byte * restrict gp)
 {
 	int k;
 	int vi = v >> 16;
@@ -211,30 +231,35 @@ template_affine_alpha_N_near_fb0(byte * restrict dp, int da, const byte * restri
 		if (ui >= 0 && ui < sw)
 		{
 			const byte *sample = sp + (ui * (sn1+sa));
-			int a = (sa ? fz_mul255(sample[sn1], alpha) : alpha);
-			if (a != 0)
+			int a = sa ? sample[sn1] : 255;
+			int aa = (sa ? fz_mul255(a, alpha) : alpha);
+			if (aa != 0)
 			{
-				int t = 255 - a;
+				int t = 255 - aa;
 				for (k = 0; k < sn1; k++)
 					dp[k] = fz_mul255(sample[k], alpha) + fz_mul255(dp[k], t);
 				for (; k < dn1; k++)
 					dp[k] = 0;
 				if (da)
-					dp[dn1] = a + fz_mul255(dp[dn1], t);
+					dp[dn1] = aa + fz_mul255(dp[dn1], t);
 				if (hp)
-					hp[0] = a + fz_mul255(hp[0], t);
+					hp[0] = a + fz_mul255(hp[0], 255 - a);
+				if (gp)
+					gp[0] = aa + fz_mul255(gp[0], t);
 			}
 		}
 		dp += dn1+da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 	}
 	while (--w);
 }
 
 static inline void
-template_affine_alpha_N_near(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, byte * restrict hp)
+template_affine_alpha_N_near(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, byte * restrict hp, byte * restrict gp)
 {
 	int k;
 
@@ -245,23 +270,28 @@ template_affine_alpha_N_near(byte * restrict dp, int da, const byte * restrict s
 		if (ui >= 0 && ui < sw && vi >= 0 && vi < sh)
 		{
 			const byte *sample = sp + (vi * ss) + (ui * (sn1+sa));
-			int a = (sa ? fz_mul255(sample[sn1], alpha) : alpha);
-			if (a != 0)
+			int a = sa ? sample[sn1] : 255;
+			int aa = (sa ? fz_mul255(a, alpha) : alpha);
+			if (aa != 0)
 			{
-				int t = 255 - a;
+				int t = 255 - aa;
 				for (k = 0; k < sn1; k++)
 					dp[k] = fz_mul255(sample[k], alpha) + fz_mul255(dp[k], t);
 				for (; k < dn1; k++)
 					dp[k] = 0;
 				if (da)
-					dp[dn1] = a + fz_mul255(dp[dn1], t);
+					dp[dn1] = aa + fz_mul255(dp[dn1], t);
 				if (hp)
-					hp[0] = a + fz_mul255(hp[0], t);
+					hp[0] = a + fz_mul255(hp[0], 255 - a);
+				if (gp)
+					gp[0] = aa + fz_mul255(gp[0], t);
 			}
 		}
 		dp += dn1+da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -269,7 +299,7 @@ template_affine_alpha_N_near(byte * restrict dp, int da, const byte * restrict s
 }
 
 static inline void
-template_affine_alpha_N_near_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, byte * restrict hp, const fz_overprint * restrict eop)
+template_affine_alpha_N_near_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	int k;
 
@@ -280,10 +310,11 @@ template_affine_alpha_N_near_op(byte * restrict dp, int da, const byte * restric
 		if (ui >= 0 && ui < sw && vi >= 0 && vi < sh)
 		{
 			const byte *sample = sp + (vi * ss) + (ui * (sn1+sa));
-			int a = (sa ? fz_mul255(sample[sn1], alpha) : alpha);
-			if (a != 0)
+			int a = sa ? sample[sn1] : 255;
+			int aa = (sa ? fz_mul255(a, alpha) : alpha);
+			if (aa != 0)
 			{
-				int t = 255 - a;
+				int t = 255 - aa;
 				for (k = 0; k < sn1; k++)
 					if (fz_overprint_component(eop, k))
 						dp[k] = fz_mul255(sample[k], alpha) + fz_mul255(dp[k], t);
@@ -291,14 +322,18 @@ template_affine_alpha_N_near_op(byte * restrict dp, int da, const byte * restric
 					if (fz_overprint_component(eop, k))
 						dp[k] = 0;
 				if (da)
-					dp[dn1] = a + fz_mul255(dp[dn1], t);
+					dp[dn1] = aa + fz_mul255(dp[dn1], t);
 				if (hp)
-					hp[0] = a + fz_mul255(hp[0], t);
+					hp[0] = a + fz_mul255(hp[0], 255 - a);
+				if (gp)
+					gp[0] = aa + fz_mul255(gp[0], t);
 			}
 		}
 		dp += dn1+da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -306,7 +341,7 @@ template_affine_alpha_N_near_op(byte * restrict dp, int da, const byte * restric
 }
 
 static inline void
-template_affine_alpha_g2rgb_near_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int alpha, byte * restrict hp)
+template_affine_alpha_g2rgb_near_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int alpha, byte * restrict hp, byte * restrict gp)
 {
 	int ui = u >> 16;
 	if (ui < 0 || ui >= sw)
@@ -319,29 +354,34 @@ template_affine_alpha_g2rgb_near_fa0(byte * restrict dp, int da, const byte * re
 		{
 			const byte *sample = sp + (vi * ss);
 			int x = fz_mul255(sample[0], alpha);
-			int a = (sa ? fz_mul255(sample[1], alpha) : alpha);
-			if (a != 0)
+			int a = sa ? sample[1] : 255;
+			int aa = (sa ? fz_mul255(a, alpha) : alpha);
+			if (aa != 0)
 			{
-				int t = 255 - a;
+				int t = 255 - aa;
 				dp[0] = x + fz_mul255(dp[0], t);
 				dp[1] = x + fz_mul255(dp[1], t);
 				dp[2] = x + fz_mul255(dp[2], t);
 				if (da)
-					dp[3] = a + fz_mul255(dp[3], t);
+					dp[3] = aa + fz_mul255(dp[3], t);
 				if (hp)
-					hp[0] = a + fz_mul255(hp[0], t);
+					hp[0] = a + fz_mul255(hp[0], 255 - a);
+				if (gp)
+					gp[0] = aa + fz_mul255(gp[0], t);
 			}
 		}
 		dp += 3 + da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		v += fb;
 	}
 	while (--w);
 }
 
 static inline void
-template_affine_alpha_g2rgb_near_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int alpha, byte * restrict hp)
+template_affine_alpha_g2rgb_near_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int alpha, byte * restrict hp, byte * restrict gp)
 {
 	int vi = v >> 16;
 	if (vi < 0 || vi >= sh)
@@ -354,29 +394,34 @@ template_affine_alpha_g2rgb_near_fb0(byte * restrict dp, int da, const byte * re
 		{
 			const byte *sample = sp + (ui * (1+sa));
 			int x = fz_mul255(sample[0], alpha);
-			int a = (sa ? fz_mul255(sample[1], alpha) : alpha);
-			if (a != 0)
+			int a = sa ? sample[1] : 255;
+			int aa = (sa ? fz_mul255(a, alpha) : alpha);
+			if (aa != 0)
 			{
-				int t = 255 - a;
+				int t = 255 - aa;
 				dp[0] = x + fz_mul255(dp[0], t);
 				dp[1] = x + fz_mul255(dp[1], t);
 				dp[2] = x + fz_mul255(dp[2], t);
 				if (da)
-					dp[3] = a + fz_mul255(dp[3], t);
+					dp[3] = aa + fz_mul255(dp[3], t);
 				if (hp)
-					hp[0] = a + fz_mul255(hp[0], t);
+					hp[0] = a + fz_mul255(hp[0], 255 - a);
+				if (gp)
+					gp[0] = aa + fz_mul255(gp[0], t);
 			}
 		}
 		dp += 3 + da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 	}
 	while (--w);
 }
 
 static inline void
-template_affine_alpha_g2rgb_near(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int alpha, byte * restrict hp)
+template_affine_alpha_g2rgb_near(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int alpha, byte * restrict hp, byte * restrict gp)
 {
 	do
 	{
@@ -386,22 +431,27 @@ template_affine_alpha_g2rgb_near(byte * restrict dp, int da, const byte * restri
 		{
 			const byte *sample = sp + (vi * ss) + (ui * (1+sa));
 			int x = fz_mul255(sample[0], alpha);
-			int a = (sa ? fz_mul255(sample[1], alpha): alpha);
-			if (a != 0)
+			int a = sa ? sample[1] : 255;
+			int aa = (sa ? fz_mul255(a, alpha): alpha);
+			if (aa != 0)
 			{
-				int t = 255 - a;
+				int t = 255 - aa;
 				dp[0] = x + fz_mul255(dp[0], t);
 				dp[1] = x + fz_mul255(dp[1], t);
 				dp[2] = x + fz_mul255(dp[2], t);
 				if (da)
-					dp[3] = a + fz_mul255(dp[3], t);
+					dp[3] = aa + fz_mul255(dp[3], t);
 				if (hp)
-					hp[0] = a + fz_mul255(hp[0], t);
+					hp[0] = a + fz_mul255(hp[0], 255 - a);
+				if (gp)
+					gp[0] = aa + fz_mul255(gp[0], t);
 			}
 		}
 		dp += 3 + da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -410,7 +460,7 @@ template_affine_alpha_g2rgb_near(byte * restrict dp, int da, const byte * restri
 
 /* Blend premultiplied source image over destination */
 static inline void
-template_affine_N_lerp(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, byte * restrict hp)
+template_affine_N_lerp(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, byte * restrict hp, byte * restrict gp)
 {
 	int k;
 
@@ -441,11 +491,15 @@ template_affine_N_lerp(byte * restrict dp, int da, const byte * restrict sp, int
 					dp[dn1] = y + fz_mul255(dp[dn1], t);
 				if (hp)
 					hp[0] = y + fz_mul255(hp[0], t);
+				if (gp)
+					gp[0] = y + fz_mul255(gp[0], t);
 			}
 		}
 		dp += dn1 + da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -453,7 +507,7 @@ template_affine_N_lerp(byte * restrict dp, int da, const byte * restrict sp, int
 }
 
 static inline void
-template_affine_N_lerp_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, byte * restrict hp, const fz_overprint * restrict eop)
+template_affine_N_lerp_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	int k;
 
@@ -486,11 +540,15 @@ template_affine_N_lerp_op(byte * restrict dp, int da, const byte * restrict sp,
 					dp[dn1] = y + fz_mul255(dp[dn1], t);
 				if (hp)
 					hp[0] = y + fz_mul255(hp[0], t);
+				if (gp)
+					gp[0] = y + fz_mul255(gp[0], t);
 			}
 		}
 		dp += dn1 + da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -498,7 +556,7 @@ template_affine_N_lerp_op(byte * restrict dp, int da, const byte * restrict sp,
 }
 
 static inline void
-template_affine_solid_g2rgb_lerp(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, byte * restrict hp)
+template_affine_solid_g2rgb_lerp(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, byte * restrict hp, byte * restrict gp)
 {
 	do
 	{
@@ -524,11 +582,15 @@ template_affine_solid_g2rgb_lerp(byte * restrict dp, int da, const byte * restri
 					dp[3] = y + fz_mul255(dp[3], t);
 				if (hp)
 					hp[0] = y + fz_mul255(hp[0], t);
+				if (gp)
+					gp[0] = y + fz_mul255(gp[0], t);
 			}
 		}
 		dp += 3 + da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -536,7 +598,7 @@ template_affine_solid_g2rgb_lerp(byte * restrict dp, int da, const byte * restri
 }
 
 static inline void
-template_affine_N_near_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, byte * restrict hp)
+template_affine_N_near_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, byte * restrict hp, byte * restrict gp)
 {
 	int k;
 	int ui = u >> 16;
@@ -576,6 +638,8 @@ template_affine_N_near_fa0(byte * restrict dp, int da, const byte * restrict sp,
 					}
 					if (hp)
 						hp[0] = a;
+					if (gp)
+						gp[0] = a;
 				}
 				else
 				{
@@ -587,19 +651,23 @@ template_affine_N_near_fa0(byte * restrict dp, int da, const byte * restrict sp,
 						dp[dn1] = a + fz_mul255(dp[dn1], t);
 					if (hp)
 						hp[0] = a + fz_mul255(hp[0], t);
+					if (gp)
+						gp[0] = a + fz_mul255(gp[0], t);
 				}
 			}
 		}
 		dp += dn1+da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		v += fb;
 	}
 	while (--w);
 }
 
 static inline void
-template_affine_N_near_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, byte * restrict hp)
+template_affine_N_near_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, byte * restrict hp, byte * restrict gp)
 {
 	int k;
 	int vi = v >> 16;
@@ -639,6 +707,8 @@ template_affine_N_near_fb0(byte * restrict dp, int da, const byte * restrict sp,
 					}
 					if (hp)
 						hp[0] = a;
+					if (gp)
+						gp[0] = a;
 				}
 				else
 				{
@@ -650,19 +720,23 @@ template_affine_N_near_fb0(byte * restrict dp, int da, const byte * restrict sp,
 						dp[dn1] = a + fz_mul255(dp[dn1], t);
 					if (hp)
 						hp[0] = a + fz_mul255(hp[0], t);
+					if (gp)
+						gp[0] = a + fz_mul255(gp[0], t);
 				}
 			}
 		}
 		dp += dn1+da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 	}
 	while (--w);
 }
 
 static inline void
-template_affine_N_near(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, byte * restrict hp)
+template_affine_N_near(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, byte * restrict hp, byte * restrict gp)
 {
 	int k;
 
@@ -700,6 +774,8 @@ template_affine_N_near(byte * restrict dp, int da, const byte * restrict sp, int
 					}
 					if (hp)
 						hp[0] = a;
+					if (gp)
+						gp[0] = a;
 				}
 				else
 				{
@@ -711,12 +787,16 @@ template_affine_N_near(byte * restrict dp, int da, const byte * restrict sp, int
 						dp[dn1] = a + fz_mul255(dp[dn1], t);
 					if (hp)
 						hp[0] = a + fz_mul255(hp[0], t);
+					if (gp)
+						gp[0] = a + fz_mul255(gp[0], t);
 				}
 			}
 		}
 		dp += dn1+da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -724,7 +804,7 @@ template_affine_N_near(byte * restrict dp, int da, const byte * restrict sp, int
 }
 
 static inline void
-template_affine_N_near_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, byte * restrict hp, const fz_overprint * restrict eop)
+template_affine_N_near_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	int k;
 
@@ -760,6 +840,8 @@ template_affine_N_near_op(byte * restrict dp, int da, const byte * restrict sp,
 						dp[dn1] = a;
 					if (hp)
 						hp[0] = a;
+					if (gp)
+						gp[0] = a;
 				}
 				else
 				{
@@ -773,12 +855,16 @@ template_affine_N_near_op(byte * restrict dp, int da, const byte * restrict sp,
 						dp[dn1] = a + fz_mul255(dp[dn1], t);
 					if (hp)
 						hp[0] = a + fz_mul255(hp[0], t);
+					if (gp)
+						gp[0] = a + fz_mul255(gp[0], t);
 				}
 			}
 		}
 		dp += dn1+da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -786,7 +872,7 @@ template_affine_N_near_op(byte * restrict dp, int da, const byte * restrict sp,
 }
 
 static inline void
-template_affine_solid_g2rgb_near_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, byte * restrict hp)
+template_affine_solid_g2rgb_near_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, byte * restrict hp, byte * restrict gp)
 {
 	int ui = u >> 16;
 	if (ui < 0 || ui >= sw)
@@ -812,6 +898,8 @@ template_affine_solid_g2rgb_near_fa0(byte * restrict dp, int da, const byte * re
 						dp[3] = a;
 					if (hp)
 						hp[0] = a;
+					if (gp)
+						gp[0] = a;
 				}
 				else
 				{
@@ -822,19 +910,23 @@ template_affine_solid_g2rgb_near_fa0(byte * restrict dp, int da, const byte * re
 						dp[3] = a + fz_mul255(dp[3], t);
 					if (hp)
 						hp[0] = a + fz_mul255(hp[0], t);
+					if (gp)
+						gp[0] = a + fz_mul255(gp[0], t);
 				}
 			}
 		}
 		dp += 3 + da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		v += fb;
 	}
 	while (--w);
 }
 
 static inline void
-template_affine_solid_g2rgb_near_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, byte * restrict hp)
+template_affine_solid_g2rgb_near_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, byte * restrict hp, byte * restrict gp)
 {
 	int vi = v >> 16;
 	if (vi < 0 || vi >= sh)
@@ -860,6 +952,8 @@ template_affine_solid_g2rgb_near_fb0(byte * restrict dp, int da, const byte * re
 						dp[3] = a;
 					if (hp)
 						hp[0] = a;
+					if (gp)
+						gp[0] = a;
 				}
 				else
 				{
@@ -870,19 +964,23 @@ template_affine_solid_g2rgb_near_fb0(byte * restrict dp, int da, const byte * re
 						dp[3] = a + fz_mul255(dp[3], t);
 					if (hp)
 						hp[0] = a + fz_mul255(hp[0], t);
+					if (gp)
+						gp[0] = a + fz_mul255(gp[0], t);
 				}
 			}
 		}
 		dp += 3 + da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 	}
 	while (--w);
 }
 
 static inline void
-template_affine_solid_g2rgb_near(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, byte * restrict hp)
+template_affine_solid_g2rgb_near(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, byte * restrict hp, byte * restrict gp)
 {
 	do
 	{
@@ -905,6 +1003,8 @@ template_affine_solid_g2rgb_near(byte * restrict dp, int da, const byte * restri
 						dp[3] = a;
 					if (hp)
 						hp[0] = a;
+					if (gp)
+						gp[0] = a;
 				}
 				else
 				{
@@ -915,12 +1015,16 @@ template_affine_solid_g2rgb_near(byte * restrict dp, int da, const byte * restri
 						dp[3] = a + fz_mul255(dp[3], t);
 					if (hp)
 						hp[0] = a + fz_mul255(hp[0], t);
+					if (gp)
+						gp[0] = a + fz_mul255(gp[0], t);
 				}
 			}
 		}
 		dp += 3 + da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -930,7 +1034,7 @@ template_affine_solid_g2rgb_near(byte * restrict dp, int da, const byte * restri
 /* Blend non-premultiplied color in source image mask over destination */
 
 static inline void
-template_affine_color_N_lerp(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int u, int v, int fa, int fb, int w, int dn1, int sn1, const byte * restrict color, byte * restrict hp)
+template_affine_color_N_lerp(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int u, int v, int fa, int fb, int w, int dn1, int sn1, const byte * restrict color, byte * restrict hp, byte * restrict gp)
 {
 	int sa = color[dn1];
 	int k;
@@ -956,12 +1060,16 @@ template_affine_color_N_lerp(byte * restrict dp, int da, const byte * restrict s
 				if (da)
 					dp[dn1] = FZ_BLEND(255, dp[dn1], masa);
 				if (hp)
-					hp[0] = FZ_BLEND(255, hp[0], masa);
+					hp[0] = FZ_BLEND(255, hp[0], ma);
+				if (gp)
+					gp[0] = FZ_BLEND(255, gp[0], masa);
 			}
 		}
 		dp += dn1 + da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -969,7 +1077,7 @@ template_affine_color_N_lerp(byte * restrict dp, int da, const byte * restrict s
 }
 
 static inline void
-template_affine_color_N_lerp_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int u, int v, int fa, int fb, int w, int dn1, int sn1, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+template_affine_color_N_lerp_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int u, int v, int fa, int fb, int w, int dn1, int sn1, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	int sa = color[dn1];
 	int k;
@@ -996,12 +1104,16 @@ template_affine_color_N_lerp_op(byte * restrict dp, int da, const byte * restric
 				if (da)
 					dp[dn1] = FZ_BLEND(255, dp[dn1], masa);
 				if (hp)
-					hp[0] = FZ_BLEND(255, hp[0], masa);
+					hp[0] = FZ_BLEND(255, hp[0], ma);
+				if (gp)
+					gp[0] = FZ_BLEND(255, gp[0], masa);
 			}
 		}
 		dp += dn1 + da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -1009,7 +1121,7 @@ template_affine_color_N_lerp_op(byte * restrict dp, int da, const byte * restric
 }
 
 static inline void
-template_affine_color_N_near(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int u, int v, int fa, int fb, int w, int dn1, int sn1, const byte * restrict color, byte * restrict hp)
+template_affine_color_N_near(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int u, int v, int fa, int fb, int w, int dn1, int sn1, const byte * restrict color, byte * restrict hp, byte * restrict gp)
 {
 	int sa = color[dn1];
 	int k;
@@ -1029,12 +1141,16 @@ template_affine_color_N_near(byte * restrict dp, int da, const byte * restrict s
 				if (da)
 					dp[dn1] = FZ_BLEND(255, dp[dn1], masa);
 				if (hp)
-					hp[0] = FZ_BLEND(255, hp[0], masa);
+					hp[0] = FZ_BLEND(255, hp[0], ma);
+				if (gp)
+					gp[0] = FZ_BLEND(255, gp[0], masa);
 			}
 		}
 		dp += dn1+da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -1042,7 +1158,7 @@ template_affine_color_N_near(byte * restrict dp, int da, const byte * restrict s
 }
 
 static inline void
-template_affine_color_N_near_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int u, int v, int fa, int fb, int w, int dn1, int sn1, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+template_affine_color_N_near_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int u, int v, int fa, int fb, int w, int dn1, int sn1, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	int sa = color[dn1];
 	int k;
@@ -1063,12 +1179,16 @@ template_affine_color_N_near_op(byte * restrict dp, int da, const byte * restric
 				if (da)
 					dp[dn1] = FZ_BLEND(255, dp[dn1], masa);
 				if (hp)
-					hp[0] = FZ_BLEND(255, hp[0], masa);
+					hp[0] = FZ_BLEND(255, hp[0], ma);
+				if (gp)
+					gp[0] = FZ_BLEND(255, gp[0], masa);
 			}
 		}
 		dp += dn1+da;
 		if (hp)
 			hp++;
+		if (gp)
+			gp++;
 		u += fa;
 		v += fb;
 	}
@@ -1076,278 +1196,278 @@ template_affine_color_N_near_op(byte * restrict dp, int da, const byte * restric
 }
 
 static void
-paint_affine_lerp_da_sa_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_sa_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, hp);
+	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_sa_alpha_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_sa_alpha_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, hp);
+	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_alpha_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_alpha_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_lerp_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 #if FZ_PLOTTERS_G
 static void
-paint_affine_lerp_da_sa_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_sa_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_sa_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_sa_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_sa_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_sa_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_lerp_sa_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_sa_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 #endif /* FZ_PLOTTERS_G */
 
 #if FZ_PLOTTERS_RGB
 static void
-paint_affine_lerp_da_sa_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_sa_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_sa_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_sa_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_sa_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_sa_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_lerp_sa_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_sa_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_lerp_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 #endif /* FZ_PLOTTERS_RGB */
 
 #if FZ_PLOTTERS_CMYK
 static void
-paint_affine_lerp_da_sa_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_sa_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_sa_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_sa_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_sa_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_sa_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_lerp_sa_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_sa_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_lerp_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 #endif /* FZ_PLOTTERS_CMYK */
 
 #if FZ_PLOTTERS_N
 static void
-paint_affine_lerp_da_sa_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_sa_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_sa_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_sa_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_sa_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_sa_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_lerp_sa_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_sa_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_lerp_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 #endif /* FZ_PLOTTERS_N */
 
 #ifdef FZ_ENABLE_SPOT_RENDERING
 static void
-paint_affine_lerp_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_lerp_op(dp, da, sp, sw, sh, ss, sa, u, v, fa, fb, w, dn, sn, hp, eop);
+	template_affine_N_lerp_op(dp, da, sp, sw, sh, ss, sa, u, v, fa, fb, w, dn, sn, hp, gp, eop);
 }
 
 static void
-paint_affine_lerp_alpha_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_alpha_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_lerp_op(dp, da, sp, sw, sh, ss, sa, u, v, fa, fb, w, dn, sn, alpha, hp, eop);
+	template_affine_alpha_N_lerp_op(dp, da, sp, sw, sh, ss, sa, u, v, fa, fb, w, dn, sn, alpha, hp, gp, eop);
 }
 #endif /* FZ_ENABLE_SPOT_RENDERING */
 
@@ -1604,59 +1724,59 @@ fz_paint_affine_lerp_spots(int da, int sa, int fa, int fb, int dn, int sn, int a
 
 #if FZ_PLOTTERS_RGB
 static void
-paint_affine_lerp_da_sa_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_sa_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_sa_g2rgb_alpha(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_sa_g2rgb_alpha(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_lerp(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_lerp_da_g2rgb_alpha(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_da_g2rgb_alpha(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_lerp(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_sa_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_sa_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_lerp_sa_g2rgb_alpha(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_sa_g2rgb_alpha(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_lerp(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_lerp_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_lerp_g2rgb_alpha(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_lerp_g2rgb_alpha(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_lerp(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static paintfn_t *
@@ -1701,782 +1821,782 @@ fz_paint_affine_g2rgb_lerp(int da, int sa, int fa, int fb, int n, int alpha)
 #endif /* FZ_PLOTTERS_RGB */
 
 static void
-paint_affine_near_da_sa_0_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_0_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, hp);
+	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_0_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_0_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_0_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_0_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, hp);
+	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_0_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_0_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_0_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_0_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, hp);
+	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_0_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_0_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_0_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_0_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, hp);
+	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_0_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_0_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, hp);
+	template_affine_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, alpha, hp);
+	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 0, 0, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, hp);
+	template_affine_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, alpha, hp);
+	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 0, 0, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int snn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int snn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_near_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_near_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 #if FZ_PLOTTERS_G
 static void
-paint_affine_near_da_sa_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_1_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_near_sa_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_1_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp);
+	template_affine_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp);
+	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 1, 1, alpha, hp, gp);
 }
 #endif /* FZ_PLOTTERS_G */
 
 #if FZ_PLOTTERS_RGB
 static void
-paint_affine_near_da_sa_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_3_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_near_sa_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_3_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_near_sa_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp);
+	template_affine_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp);
+	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 3, 3, alpha, hp, gp);
 }
 #endif /* FZ_PLOTTERS_RGB */
 
 #if FZ_PLOTTERS_CMYK
 static void
-paint_affine_near_da_sa_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_4_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_4_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp);
+	template_affine_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp);
+	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, 4, 4, alpha, hp, gp);
 }
 #endif /* FZ_PLOTTERS_CMYK */
 
 #if FZ_PLOTTERS_N
 static void
-paint_affine_near_da_sa_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_N_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_N_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp);
+	template_affine_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp);
+	template_affine_alpha_N_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, dn, sn, alpha, hp, gp);
 }
 #endif /* FZ_PLOTTERS_N */
 
 #ifdef FZ_ENABLE_SPOT_RENDERING
 static void
-paint_affine_near_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_N_near_op(dp, da, sp, sw, sh, ss, sa, u, v, fa, fb, w, dn, sn, hp, eop);
+	template_affine_N_near_op(dp, da, sp, sw, sh, ss, sa, u, v, fa, fb, w, dn, sn, hp, gp, eop);
 }
 
 static void
-paint_affine_near_alpha_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_N_near_op(dp, da, sp, sw, sh, ss, sa, u, v, fa, fb, w, dn, sn, alpha, hp, eop);
+	template_affine_alpha_N_near_op(dp, da, sp, sw, sh, ss, sa, u, v, fa, fb, w, dn, sn, alpha, hp, gp, eop);
 }
 #endif /* FZ_ENABLE_SPOT_RENDERING */
 
@@ -3039,171 +3159,171 @@ fz_paint_affine_near_spots(int da, int sa, int fa, int fb, int dn, int sn, int a
 
 #if FZ_PLOTTERS_RGB
 static void
-paint_affine_near_da_sa_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_near_fa0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_near_fa0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_near_fa0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_g2rgb_fa0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_near_fa0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_near_fb0(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_near_fb0(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_near_fb0(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_g2rgb_fb0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_near_fb0(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_near_da_sa_alpha_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_sa_alpha_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_near(dp, 1, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_da_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_near_da_alpha_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_da_alpha_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_near(dp, 1, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_sa_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_near_sa_alpha_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_sa_alpha_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_near(dp, 0, sp, sw, sh, ss, 1, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static void
-paint_affine_near_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_solid_g2rgb_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp);
+	template_affine_solid_g2rgb_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, hp, gp);
 }
 
 static void
-paint_affine_near_alpha_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_near_alpha_g2rgb(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn1, int sn1, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_alpha_g2rgb_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp);
+	template_affine_alpha_g2rgb_near(dp, 0, sp, sw, sh, ss, 0, u, v, fa, fb, w, alpha, hp, gp);
 }
 
 static paintfn_t *
@@ -3316,82 +3436,82 @@ fz_paint_affine_g2rgb_near(int da, int sa, int fa, int fb, int n, int alpha)
 #endif /* FZ_PLOTTERS_RGB */
 
 static void
-paint_affine_color_lerp_da_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_lerp_da_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_lerp(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 0, 0, color, hp);
+	template_affine_color_N_lerp(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 0, 0, color, hp, gp);
 }
 
 #if FZ_PLOTTERS_G
 static void
-paint_affine_color_lerp_da_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_lerp_da_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_lerp(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 1, 1, color, hp);
+	template_affine_color_N_lerp(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 1, 1, color, hp, gp);
 }
 
 static void
-paint_affine_color_lerp_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_lerp_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_lerp(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, 1, 1, color, hp);
+	template_affine_color_N_lerp(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, 1, 1, color, hp, gp);
 }
 #endif /* FZ_PLOTTERS_G */
 
 #if FZ_PLOTTERS_RGB
 static void
-paint_affine_color_lerp_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_lerp_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_lerp(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, 3, 3, color, hp);
+	template_affine_color_N_lerp(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, 3, 3, color, hp, gp);
 }
 
 static void
-paint_affine_color_lerp_da_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_lerp_da_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_lerp(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 3, 3, color, hp);
+	template_affine_color_N_lerp(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 3, 3, color, hp, gp);
 }
 #endif /* FZ_PLOTTERS_RGB */
 
 #if FZ_PLOTTERS_CMYK
 static void
-paint_affine_color_lerp_da_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_lerp_da_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_lerp(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 4, 4, color, hp);
+	template_affine_color_N_lerp(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 4, 4, color, hp, gp);
 }
 
 static void
-paint_affine_color_lerp_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_lerp_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_lerp(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, 4, 4, color, hp);
+	template_affine_color_N_lerp(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, 4, 4, color, hp, gp);
 }
 #endif /* FZ_PLOTTERS_G */
 
 #if FZ_PLOTTERS_N
 static void
-paint_affine_color_lerp_da_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_lerp_da_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_lerp(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp);
+	template_affine_color_N_lerp(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp, gp);
 }
 
 static void
-paint_affine_color_lerp_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_lerp_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_lerp(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp);
+	template_affine_color_N_lerp(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp, gp);
 }
 #endif /* FZ_PLOTTERS_N */
 
 #ifdef FZ_ENABLE_SPOT_RENDERING
 static void
-paint_affine_color_lerp_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_lerp_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_lerp_op(dp, da, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp, eop);
+	template_affine_color_N_lerp_op(dp, da, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp, gp, eop);
 }
 #endif /* FZ_ENABLE_SPOT_RENDERING */
 
@@ -3432,89 +3552,89 @@ fz_paint_affine_color_lerp_spots(int da, int sa, int fa, int fb, int dn, int sn,
 #endif /* FZ_ENABLE_SPOT_RENDERING */
 
 static void
-paint_affine_color_near_da_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_near_da_0(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_near(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 0, 0, color, hp);
+	template_affine_color_N_near(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 0, 0, color, hp, gp);
 }
 
 #if FZ_PLOTTERS_G
 static void
-paint_affine_color_near_da_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_near_da_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_near(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 1, 1, color, hp);
+	template_affine_color_N_near(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 1, 1, color, hp, gp);
 }
 
 static void
-paint_affine_color_near_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_near_1(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_near(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, 1, 1, color, hp);
+	template_affine_color_N_near(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, 1, 1, color, hp, gp);
 }
 #endif /* FZ_PLOTTERS_G */
 
 #if FZ_PLOTTERS_RGB
 static void
-paint_affine_color_near_da_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_near_da_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_near(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 3, 3, color, hp);
+	template_affine_color_N_near(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 3, 3, color, hp, gp);
 }
 
 static void
-paint_affine_color_near_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_near_3(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_near(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, 3, 3, color, hp);
+	template_affine_color_N_near(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, 3, 3, color, hp, gp);
 }
 #endif /* FZ_PLOTTERS_RGB */
 
 #if FZ_PLOTTERS_CMYK
 static void
-paint_affine_color_near_da_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_near_da_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_near(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 4, 4, color, hp);
+	template_affine_color_N_near(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, 4, 4, color, hp, gp);
 }
 
 static void
-paint_affine_color_near_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_near_4(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_near(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, 4, 4, color, hp);
+	template_affine_color_N_near(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, 4, 4, color, hp, gp);
 }
 #endif /* FZ_PLOTTERS_CMYK */
 
 #if FZ_PLOTTERS_N
 static void
-paint_affine_color_near_da_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_near_da_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_near(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp);
+	template_affine_color_N_near(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp, gp);
 }
 
 static void
-paint_affine_color_near_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_near_N(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_near(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp);
+	template_affine_color_N_near(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp, gp);
 }
 #endif /* FZ_PLOTTERS_N */
 
 #ifdef FZ_ENABLE_SPOT_RENDERING
 static void
-paint_affine_color_near_da_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_near_da_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_near_op(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp, eop);
+	template_affine_color_N_near_op(dp, 1, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp, gp, eop);
 }
 
 static void
-paint_affine_color_near_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, const fz_overprint * restrict eop)
+paint_affine_color_near_N_op(byte * restrict dp, int da, const byte * restrict sp, int sw, int sh, int ss, int sa, int u, int v, int fa, int fb, int w, int dn, int sn, int alpha, const byte * restrict color, byte * restrict hp, byte * restrict gp, const fz_overprint * restrict eop)
 {
 	TRACK_FN();
-	template_affine_color_N_near_op(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp, eop);
+	template_affine_color_N_near_op(dp, 0, sp, sw, sh, ss, u, v, fa, fb, w, dn, sn, color, hp, gp, eop);
 }
 #endif /* FZ_ENABLE_SPOT_RENDERING */
 
@@ -3747,12 +3867,12 @@ fz_gridfit_matrix(int as_tiled, fz_matrix *m)
 /* Draw an image with an affine transform on destination */
 
 static void
-fz_paint_image_imp(fz_pixmap * restrict dst, const fz_irect *scissor, const fz_pixmap * restrict shape, const fz_pixmap * restrict img, const fz_matrix * restrict ctm, const byte * restrict color, int alpha, int lerp_allowed, int as_tiled, const fz_overprint * restrict eop)
+fz_paint_image_imp(fz_pixmap * restrict dst, const fz_irect *scissor, fz_pixmap * restrict shape, fz_pixmap * restrict group_alpha, const fz_pixmap * restrict img, const fz_matrix * restrict ctm, const byte * restrict color, int alpha, int lerp_allowed, int as_tiled, const fz_overprint * restrict eop)
 {
-	byte *dp, *sp, *hp;
+	byte *dp, *sp, *hp, *gp;
 	int u, v, fa, fb, fc, fd;
 	int x, y, w, h;
-	int sw, sh, ss, sa, sn, hs, da, dn;
+	int sw, sh, ss, sa, sn, hs, da, dn, gs;
 	fz_irect bbox;
 	int dolerp;
 	paintfn_t *paintfn;
@@ -3792,16 +3912,24 @@ fz_paint_image_imp(fz_pixmap * restrict dst, const fz_irect *scissor, const fz_p
 	x = bbox.x0;
 	if (shape && shape->x > x)
 		x = shape->x;
+	if (group_alpha && group_alpha->x > x)
+		x = group_alpha->x;
 	y = bbox.y0;
 	if (shape && shape->y > y)
 		y = shape->y;
+	if (group_alpha && group_alpha->y > y)
+		y = group_alpha->y;
 	w = bbox.x1;
 	if (shape && shape->x + shape->w < w)
 		w = shape->x + shape->w;
+	if (group_alpha && group_alpha->x + group_alpha->w < w)
+		w = group_alpha->x + group_alpha->w;
 	w -= x;
 	h = bbox.y1;
 	if (shape && shape->y + shape->h < h)
 		h = shape->y + shape->h;
+	if (group_alpha && group_alpha->y + group_alpha->h < h)
+		h = group_alpha->y + group_alpha->h;
 	h -= y;
 	if (w <= 0 || h <= 0)
 		return;
@@ -3842,6 +3970,16 @@ fz_paint_image_imp(fz_pixmap * restrict dst, const fz_irect *scissor, const fz_p
 		hs = 0;
 		hp = NULL;
 	}
+	if (group_alpha)
+	{
+		gs = group_alpha->stride;
+		gp = group_alpha->samples + (unsigned int)((y - group_alpha->y) * group_alpha->stride + x - group_alpha->x);
+	}
+	else
+	{
+		gs = 0;
+		gp = NULL;
+	}
 
 	/* TODO: if (fb == 0 && fa == 1) call fz_paint_span */
 
@@ -3916,23 +4054,24 @@ fz_paint_image_imp(fz_pixmap * restrict dst, const fz_irect *scissor, const fz_p
 
 	while (h--)
 	{
-		paintfn(dp, da, sp, sw, sh, ss, sa, u, v, fa, fb, w, dn, sn, alpha, color, hp, eop);
+		paintfn(dp, da, sp, sw, sh, ss, sa, u, v, fa, fb, w, dn, sn, alpha, color, hp, gp, eop);
 		dp += dst->stride;
 		hp += hs;
+		gp += gs;
 		u += fc;
 		v += fd;
 	}
 }
 
 void
-fz_paint_image_with_color(fz_pixmap * restrict dst, const fz_irect * restrict scissor, fz_pixmap * restrict shape, const fz_pixmap * restrict img, const fz_matrix * restrict ctm, const byte * restrict color, int lerp_allowed, int as_tiled, const fz_overprint * restrict eop)
+fz_paint_image_with_color(fz_pixmap * restrict dst, const fz_irect * restrict scissor, fz_pixmap * restrict shape, fz_pixmap * restrict group_alpha, const fz_pixmap * restrict img, const fz_matrix * restrict ctm, const byte * restrict color, int lerp_allowed, int as_tiled, const fz_overprint * restrict eop)
 {
 	assert(img->n == 1);
-	fz_paint_image_imp(dst, scissor, shape, img, ctm, color, 255, lerp_allowed, as_tiled, eop);
+	fz_paint_image_imp(dst, scissor, shape, group_alpha, img, ctm, color, 255, lerp_allowed, as_tiled, eop);
 }
 
 void
-fz_paint_image(fz_pixmap * restrict dst, const fz_irect * restrict scissor, fz_pixmap * restrict shape, const fz_pixmap * restrict img, const fz_matrix * restrict ctm, int alpha, int lerp_allowed, int as_tiled, const fz_overprint * restrict eop)
+fz_paint_image(fz_pixmap * restrict dst, const fz_irect * restrict scissor, fz_pixmap * restrict shape, fz_pixmap * restrict group_alpha, const fz_pixmap * restrict img, const fz_matrix * restrict ctm, int alpha, int lerp_allowed, int as_tiled, const fz_overprint * restrict eop)
 {
-	fz_paint_image_imp(dst, scissor, shape, img, ctm, NULL, alpha, lerp_allowed, as_tiled, eop);
+	fz_paint_image_imp(dst, scissor, shape, group_alpha, img, ctm, NULL, alpha, lerp_allowed, as_tiled, eop);
 }
diff --git a/source/fitz/draw-device.c b/source/fitz/draw-device.c
index 27229108..ff6097fb 100644
--- a/source/fitz/draw-device.c
+++ b/source/fitz/draw-device.c
@@ -31,6 +31,7 @@ struct fz_draw_state_s {
 	fz_pixmap *dest;
 	fz_pixmap *mask;
 	fz_pixmap *shape;
+	fz_pixmap *group_alpha;
 	int blendmode;
 	int id, encache;
 	float alpha;
@@ -198,6 +199,8 @@ static void emergency_pop_stack(fz_context *ctx, fz_draw_device *dev, fz_draw_st
 		fz_drop_pixmap(ctx, state[1].dest);
 	if (state[1].shape != state[0].shape)
 		fz_drop_pixmap(ctx, state[1].shape);
+	if (state[1].group_alpha != state[0].group_alpha)
+		fz_drop_pixmap(ctx, state[1].group_alpha);
 	dev->top--;
 	STACK_POPPED("emergency");
 	fz_rethrow(ctx);
@@ -242,13 +245,14 @@ fz_knockout_begin(fz_context *ctx, fz_draw_device *dev)
 			fz_clear_pixmap(ctx, dest);
 	}
 
+	/* Knockout groups (and only knockout groups) rely on shape */
 	shape = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
 	fz_clear_pixmap(ctx, shape);
 #ifdef DUMP_GROUP_BLENDS
 	dump_spaces(dev->top-1, "");
 	fz_dump_blend(ctx, "Knockout begin: background is ", dest);
 	if (shape)
-		fz_dump_blend(ctx, "/", shape);
+		fz_dump_blend(ctx, "/S=", shape);
 	printf("\n");
 #endif
 	state[1].scissor = bbox;
@@ -282,20 +286,22 @@ static void fz_knockout_end(fz_context *ctx, fz_draw_device *dev)
 	dump_spaces(dev->top, "");
 	fz_dump_blend(ctx, "Knockout end: blending ", state[1].dest);
 	if (state[1].shape)
-		fz_dump_blend(ctx, "/", state[1].shape);
+		fz_dump_blend(ctx, "/S=", state[1].shape);
+	if (state[1].group_alpha)
+		fz_dump_blend(ctx, "/GA=", state[1].group_alpha);
 	fz_dump_blend(ctx, " onto ", state[0].dest);
 	if (state[0].shape)
-		fz_dump_blend(ctx, "/", state[0].shape);
+		fz_dump_blend(ctx, "/S=", state[0].shape);
+	if (state[0].group_alpha)
+		fz_dump_blend(ctx, "/GA=", state[0].group_alpha);
 	if (blendmode != 0)
 		printf(" (blend %d)", blendmode);
 	if (isolated != 0)
 		printf(" (isolated)");
 	printf(" (knockout)");
 #endif
-	if ((blendmode == 0) && (state[0].shape == state[1].shape))
-		fz_paint_pixmap(state[0].dest, state[1].dest, 255);
-	else
-		fz_blend_pixmap(ctx, state[0].dest, state[1].dest, 255, blendmode, isolated, state[1].shape);
+	assert(state[1].shape);
+	fz_blend_pixmap(ctx, state[0].dest, state[1].dest, 255, blendmode, 0, state[1].shape);
 
 	/* The following test should not be required, but just occasionally
 	 * errors can cause the stack to get out of sync, and this saves our
@@ -308,10 +314,18 @@ static void fz_knockout_end(fz_context *ctx, fz_draw_device *dev)
 			fz_paint_pixmap(state[0].shape, state[1].shape, 255);
 		fz_drop_pixmap(ctx, state[1].shape);
 	}
+	if (state[0].group_alpha != state[1].group_alpha)
+	{
+		if (state[0].group_alpha)
+			fz_paint_pixmap(state[0].group_alpha, state[1].group_alpha, 255);
+		fz_drop_pixmap(ctx, state[1].group_alpha);
+	}
 #ifdef DUMP_GROUP_BLENDS
 	fz_dump_blend(ctx, " to get ", state[0].dest);
 	if (state[0].shape)
-		fz_dump_blend(ctx, "/", state[0].shape);
+		fz_dump_blend(ctx, "/S=", state[0].shape);
+	if (state[0].group_alpha)
+		fz_dump_blend(ctx, "/GA=", state[0].group_alpha);
 	printf("\n");
 #endif
 }
@@ -557,9 +571,17 @@ fz_draw_fill_path(fz_context *ctx, fz_device *devp, const fz_path *path, int eve
 		if (!rast->fns.reusable)
 			fz_flatten_fill_path(ctx, rast, path, &ctm, flatness, &bbox, NULL);
 
-		colorbv[0] = alpha * 255;
+		colorbv[0] = 255;
 		fz_convert_rasterizer(ctx, rast, even_odd, state->shape, colorbv, 0);
 	}
+	if (state->group_alpha)
+	{
+		if (!rast->fns.reusable)
+			fz_flatten_fill_path(ctx, rast, path, &ctm, flatness, &bbox, NULL);
+
+		colorbv[0] = alpha * 255;
+		fz_convert_rasterizer(ctx, rast, even_odd, state->group_alpha, colorbv, 0);
+	}
 
 	if (state->blendmode & FZ_BLEND_KNOCKOUT)
 		fz_knockout_end(ctx, dev);
@@ -607,7 +629,9 @@ fz_draw_stroke_path(fz_context *ctx, fz_device *devp, const fz_path *path, const
 	dump_spaces(dev->top, "");
 	fz_dump_blend(ctx, "Before stroke ", state->dest);
 	if (state->shape)
-		fz_dump_blend(ctx, "/", state->shape);
+		fz_dump_blend(ctx, "/S=", state->shape);
+	if (state->group_alpha)
+		fz_dump_blend(ctx, "/GA=", state->group_alpha);
 	printf("\n");
 #endif
 	fz_convert_rasterizer(ctx, rast, 0, state->dest, colorbv, eop);
@@ -619,11 +643,21 @@ fz_draw_stroke_path(fz_context *ctx, fz_device *devp, const fz_path *path, const
 		colorbv[0] = 255;
 		fz_convert_rasterizer(ctx, rast, 0, state->shape, colorbv, 0);
 	}
+	if (state->group_alpha)
+	{
+		if (!rast->fns.reusable)
+			(void)fz_flatten_stroke_path(ctx, rast, path, stroke, &ctm, flatness, linewidth, &bbox, NULL);
+
+		colorbv[0] = 255 * alpha;
+		fz_convert_rasterizer(ctx, rast, 0, state->group_alpha, colorbv, 0);
+	}
 #ifdef DUMP_GROUP_BLENDS
 	dump_spaces(dev->top, "");
 	fz_dump_blend(ctx, "After stroke ", state->dest);
 	if (state->shape)
-		fz_dump_blend(ctx, "/", state->shape);
+		fz_dump_blend(ctx, "/S=", state->shape);
+	if (state->group_alpha)
+		fz_dump_blend(ctx, "/GA=", state->group_alpha);
 	printf("\n");
 #endif
 
@@ -685,6 +719,11 @@ fz_draw_clip_path(fz_context *ctx, fz_device *devp, const fz_path *path, int eve
 			state[1].shape = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
 			fz_clear_pixmap(ctx, state[1].shape);
 		}
+		if (state[1].group_alpha)
+		{
+			state[1].group_alpha = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
+			fz_clear_pixmap(ctx, state[1].group_alpha);
+		}
 
 		fz_convert_rasterizer(ctx, rast, even_odd, state[1].mask, NULL, 0);
 
@@ -767,6 +806,11 @@ fz_draw_clip_stroke_path(fz_context *ctx, fz_device *devp, const fz_path *path,
 			state[1].shape = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
 			fz_clear_pixmap(ctx, state[1].shape);
 		}
+		if (state->group_alpha)
+		{
+			state[1].group_alpha = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
+			fz_clear_pixmap(ctx, state[1].group_alpha);
+		}
 
 		fz_convert_rasterizer(ctx, rast, 0, state[1].mask, NULL, 0);
 
@@ -860,7 +904,7 @@ fz_draw_fill_text(fz_context *ctx, fz_device *devp, const fz_text *text, const f
 	fz_draw_state *state = &dev->stack[dev->top];
 	fz_colorspace *model = state->dest->colorspace;
 	unsigned char colorbv[FZ_MAX_COLORS + 1];
-	unsigned char shapebv;
+	unsigned char shapebv, shapebva;
 	fz_text_span *span;
 	int i;
 	fz_colorspace *colorspace = NULL;
@@ -885,6 +929,7 @@ fz_draw_fill_text(fz_context *ctx, fz_device *devp, const fz_text *text, const f
 
 	eop = resolve_color(ctx, &op, color, colorspace, alpha, color_params, colorbv, state->dest);
 	shapebv = 255;
+	shapebva = 255 * alpha;
 
 	for (span = text->head; span; span = span->next)
 	{
@@ -915,13 +960,15 @@ fz_draw_fill_text(fz_context *ctx, fz_device *devp, const fz_text *text, const f
 					draw_glyph(colorbv, state->dest, glyph, x, y, &state->scissor, eop);
 					if (state->shape)
 						draw_glyph(&shapebv, state->shape, glyph, x, y, &state->scissor, 0);
+					if (state->group_alpha)
+						draw_glyph(&shapebva, state->group_alpha, glyph, x, y, &state->scissor, 0);
 				}
 				else
 				{
 					fz_matrix mat;
 					mat.a = pixmap->w; mat.b = mat.c = 0; mat.d = pixmap->h;
 					mat.e = x + pixmap->x; mat.f = y + pixmap->y;
-					fz_paint_image(state->dest, &state->scissor, state->shape, pixmap, &mat, alpha * 255, !(devp->hints & FZ_DONT_INTERPOLATE_IMAGES), devp->flags & FZ_DEVFLAG_GRIDFIT_AS_TILED, eop);
+					fz_paint_image(state->dest, &state->scissor, state->shape, state->group_alpha, pixmap, &mat, alpha * 255, !(devp->hints & FZ_DONT_INTERPOLATE_IMAGES), devp->flags & FZ_DEVFLAG_GRIDFIT_AS_TILED, eop);
 				}
 				fz_drop_glyph(ctx, glyph);
 			}
@@ -953,6 +1000,8 @@ fz_draw_stroke_text(fz_context *ctx, fz_device *devp, const fz_text *text, const
 	fz_matrix ctm = concat(in_ctm, &dev->transform);
 	fz_draw_state *state = &dev->stack[dev->top];
 	unsigned char colorbv[FZ_MAX_COLORS + 1];
+	unsigned char solid = 255;
+	unsigned char alpha_byte = alpha * 255;
 	fz_text_span *span;
 	int i;
 	fz_colorspace *colorspace = NULL;
@@ -996,7 +1045,9 @@ fz_draw_stroke_text(fz_context *ctx, fz_device *devp, const fz_text *text, const
 				int y = (int)trm.f;
 				draw_glyph(colorbv, state->dest, glyph, x, y, &state->scissor, eop);
 				if (state->shape)
-					draw_glyph(colorbv, state->shape, glyph, x, y, &state->scissor, 0);
+					draw_glyph(&solid, state->shape, glyph, x, y, &state->scissor, 0);
+				if (state->group_alpha)
+					draw_glyph(&alpha_byte, state->group_alpha, glyph, x, y, &state->scissor, 0);
 				fz_drop_glyph(ctx, glyph);
 			}
 			else
@@ -1025,7 +1076,7 @@ fz_draw_clip_text(fz_context *ctx, fz_device *devp, const fz_text *text, const f
 	fz_draw_device *dev = (fz_draw_device*)devp;
 	fz_matrix ctm = concat(in_ctm, &dev->transform);
 	fz_irect bbox;
-	fz_pixmap *mask, *dest, *shape;
+	fz_pixmap *mask, *dest, *shape, *group_alpha;
 	fz_matrix tm, trm;
 	fz_glyph *glyph;
 	int i, gid;
@@ -1073,12 +1124,20 @@ fz_draw_clip_text(fz_context *ctx, fz_device *devp, const fz_text *text, const f
 		}
 		else
 			shape = NULL;
+		if (state->group_alpha)
+		{
+			group_alpha = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
+			fz_clear_pixmap(ctx, group_alpha);
+		}
+		else
+			group_alpha = NULL;
 
 		state[1].blendmode |= FZ_BLEND_ISOLATED;
 		state[1].scissor = bbox;
 		state[1].dest = dest;
 		state[1].mask = mask;
 		state[1].shape = shape;
+		state[1].group_alpha = group_alpha;
 #ifdef DUMP_GROUP_BLENDS
 		dump_spaces(dev->top-1, "Clip (text) begin\n");
 #endif
@@ -1107,6 +1166,8 @@ fz_draw_clip_text(fz_context *ctx, fz_device *devp, const fz_text *text, const f
 						draw_glyph(NULL, mask, glyph, x, y, &bbox, 0);
 						if (state[1].shape)
 							draw_glyph(NULL, state[1].shape, glyph, x, y, &bbox, 0);
+						if (state[1].group_alpha)
+							draw_glyph(NULL, state[1].group_alpha, glyph, x, y, &bbox, 0);
 						fz_drop_glyph(ctx, glyph);
 					}
 					else
@@ -1157,7 +1218,7 @@ fz_draw_clip_stroke_text(fz_context *ctx, fz_device *devp, const fz_text *text,
 	fz_draw_device *dev = (fz_draw_device*)devp;
 	fz_matrix ctm = concat(in_ctm, &dev->transform);
 	fz_irect bbox;
-	fz_pixmap *mask, *dest, *shape;
+	fz_pixmap *mask, *dest, *shape, *group_alpha;
 	fz_matrix tm, trm;
 	fz_glyph *glyph;
 	int i, gid;
@@ -1202,6 +1263,13 @@ fz_draw_clip_stroke_text(fz_context *ctx, fz_device *devp, const fz_text *text,
 		}
 		else
 			shape = state->shape;
+		if (state->group_alpha)
+		{
+			state[1].group_alpha = group_alpha = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
+			fz_clear_pixmap(ctx, group_alpha);
+		}
+		else
+			group_alpha = NULL;
 
 		state[1].blendmode |= FZ_BLEND_ISOLATED;
 		state[1].scissor = bbox;
@@ -1233,6 +1301,8 @@ fz_draw_clip_stroke_text(fz_context *ctx, fz_device *devp, const fz_text *text,
 						draw_glyph(NULL, mask, glyph, x, y, &bbox, 0);
 						if (shape)
 							draw_glyph(NULL, shape, glyph, x, y, &bbox, 0);
+						if (group_alpha)
+							draw_glyph(NULL, group_alpha, glyph, x, y, &bbox, 0);
 						fz_drop_glyph(ctx, glyph);
 					}
 					else
@@ -1289,8 +1359,9 @@ fz_draw_fill_shade(fz_context *ctx, fz_device *devp, fz_shade *shade, const fz_m
 	fz_matrix ctm = concat(in_ctm, &dev->transform);
 	fz_rect bounds;
 	fz_irect bbox, scissor;
-	fz_pixmap *dest, *shape;
+	fz_pixmap *dest, *shape, *group_alpha;
 	unsigned char colorbv[FZ_MAX_COLORS + 1];
+	unsigned char alpha_byte = 255 * alpha;
 	fz_draw_state *state = &dev->stack[dev->top];
 	fz_overprint op = { { 0 } };
 	fz_overprint *eop;
@@ -1314,6 +1385,7 @@ fz_draw_fill_shade(fz_context *ctx, fz_device *devp, fz_shade *shade, const fz_m
 
 	dest = state->dest;
 	shape = state->shape;
+	group_alpha = state->group_alpha;
 
 	if (alpha < 1)
 	{
@@ -1327,6 +1399,11 @@ fz_draw_fill_shade(fz_context *ctx, fz_device *devp, fz_shade *shade, const fz_m
 			shape = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
 			fz_clear_pixmap(ctx, shape);
 		}
+		if (group_alpha)
+		{
+			group_alpha = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
+			fz_clear_pixmap(ctx, group_alpha);
+		}
 	}
 
 	if (shade->use_background)
@@ -1383,6 +1460,17 @@ fz_draw_fill_shade(fz_context *ctx, fz_device *devp, fz_shade *shade, const fz_m
 				}
 			}
 		}
+		if (group_alpha)
+		{
+			for (y = scissor.y0; y < scissor.y1; y++)
+			{
+				s = group_alpha->samples + (unsigned int)((y - group_alpha->y) * group_alpha->stride + (scissor.x0 - group_alpha->x));
+				for (x = scissor.x0; x < scissor.x1; x++)
+				{
+					*s++ = alpha_byte;
+				}
+			}
+		}
 	}
 
 	if (color_params->op)
@@ -1395,12 +1483,16 @@ fz_draw_fill_shade(fz_context *ctx, fz_device *devp, fz_shade *shade, const fz_m
 	fz_paint_shade(ctx, shade, colorspace, &ctm, dest, color_params, &bbox, eop);
 	if (shape)
 		fz_clear_pixmap_rect_with_value(ctx, shape, 255, &bbox);
+	if (group_alpha)
+		fz_clear_pixmap_rect_with_value(ctx, group_alpha, 255, &bbox);
 
 #ifdef DUMP_GROUP_BLENDS
 	dump_spaces(dev->top, "");
 	fz_dump_blend(ctx, "Shade ", dest);
 	if (shape)
-		fz_dump_blend(ctx, "/", shape);
+		fz_dump_blend(ctx, "/S=", shape);
+	if (group_alpha)
+		fz_dump_blend(ctx, "/GA=", group_alpha);
 	printf("\n");
 #endif
 
@@ -1411,9 +1503,14 @@ fz_draw_fill_shade(fz_context *ctx, fz_device *devp, fz_shade *shade, const fz_m
 		fz_drop_pixmap(ctx, dest);
 		if (shape)
 		{
-			fz_paint_pixmap(state->shape, shape, alpha * 255);
+			fz_paint_pixmap(state->shape, shape, 255);
 			fz_drop_pixmap(ctx, shape);
 		}
+		if (group_alpha)
+		{
+			fz_paint_pixmap(state->group_alpha, group_alpha, alpha * 255);
+			fz_drop_pixmap(ctx, group_alpha);
+		}
 	}
 
 	if (state->blendmode & FZ_BLEND_KNOCKOUT)
@@ -1650,7 +1747,7 @@ fz_draw_fill_image(fz_context *ctx, fz_device *devp, fz_image *image, const fz_m
 				pixmap = convert_pixmap_for_painting(ctx, pixmap, model, src_cs, state->dest, color_params, dev, &eop);
 		}
 
-		fz_paint_image(state->dest, &state->scissor, state->shape, pixmap, &local_ctm, alpha * 255, !(devp->hints & FZ_DONT_INTERPOLATE_IMAGES), devp->flags & FZ_DEVFLAG_GRIDFIT_AS_TILED, eop);
+		fz_paint_image(state->dest, &state->scissor, state->shape, state->group_alpha, pixmap, &local_ctm, alpha * 255, !(devp->hints & FZ_DONT_INTERPOLATE_IMAGES), devp->flags & FZ_DEVFLAG_GRIDFIT_AS_TILED, eop);
 
 		if (state->blendmode & FZ_BLEND_KNOCKOUT)
 			fz_knockout_end(ctx, dev);
@@ -1759,7 +1856,7 @@ fz_draw_fill_image_mask(fz_context *ctx, fz_device *devp, fz_image *image, const
 
 		eop = resolve_color(ctx, &op, color, colorspace, alpha, color_params, colorbv, state->dest);
 
-		fz_paint_image_with_color(state->dest, &state->scissor, state->shape, pixmap, &local_ctm, colorbv, !(devp->hints & FZ_DONT_INTERPOLATE_IMAGES), devp->flags & FZ_DEVFLAG_GRIDFIT_AS_TILED, eop);
+		fz_paint_image_with_color(state->dest, &state->scissor, state->shape, state->group_alpha, pixmap, &local_ctm, colorbv, !(devp->hints & FZ_DONT_INTERPOLATE_IMAGES), devp->flags & FZ_DEVFLAG_GRIDFIT_AS_TILED, eop);
 
 		if (state->blendmode & FZ_BLEND_KNOCKOUT)
 			fz_knockout_end(ctx, dev);
@@ -1837,6 +1934,11 @@ fz_draw_clip_image_mask(fz_context *ctx, fz_device *devp, fz_image *image, const
 			state[1].shape = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
 			fz_clear_pixmap(ctx, state[1].shape);
 		}
+		if (state[0].group_alpha)
+		{
+			state[1].group_alpha = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
+			fz_clear_pixmap(ctx, state[1].group_alpha);
+		}
 
 		state[1].blendmode |= FZ_BLEND_ISOLATED;
 		state[1].scissor = bbox;
@@ -1864,13 +1966,17 @@ fz_draw_clip_image_mask(fz_context *ctx, fz_device *devp, fz_image *image, const
 		fz_dump_blend(ctx, "Creating imagemask: plotting ", pixmap);
 		fz_dump_blend(ctx, " onto ", state[1].mask);
 		if (state[1].shape)
-			fz_dump_blend(ctx, "/", state[1].shape);
+			fz_dump_blend(ctx, "/S=", state[1].shape);
+		if (state[1].group_alpha)
+			fz_dump_blend(ctx, "/GA=", state[1].group_alpha);
 #endif
-		fz_paint_image(state[1].mask, &bbox, state[1].shape, pixmap, &local_ctm, 255, !(devp->hints & FZ_DONT_INTERPOLATE_IMAGES), devp->flags & FZ_DEVFLAG_GRIDFIT_AS_TILED, 0);
+		fz_paint_image(state[1].mask, &bbox, state[1].shape, state[1].group_alpha, pixmap, &local_ctm, 255, !(devp->hints & FZ_DONT_INTERPOLATE_IMAGES), devp->flags & FZ_DEVFLAG_GRIDFIT_AS_TILED, 0);
 #ifdef DUMP_GROUP_BLENDS
 		fz_dump_blend(ctx, " to get ", state[1].mask);
 		if (state[1].shape)
-			fz_dump_blend(ctx, "/", state[1].shape);
+			fz_dump_blend(ctx, "/S=", state[1].shape);
+		if (state[1].group_alpha)
+			fz_dump_blend(ctx, "/GA=", state[1].group_alpha);
 		printf("\n");
 #endif
 	}
@@ -1903,10 +2009,14 @@ fz_draw_pop_clip(fz_context *ctx, fz_device *devp)
 		dump_spaces(dev->top, "");
 		fz_dump_blend(ctx, "Clipping ", state[1].dest);
 		if (state[1].shape)
-			fz_dump_blend(ctx, "/", state[1].shape);
+			fz_dump_blend(ctx, "/S=", state[1].shape);
+		if (state[1].group_alpha)
+			fz_dump_blend(ctx, "/GA=", state[1].group_alpha);
 		fz_dump_blend(ctx, " onto ", state[0].dest);
 		if (state[0].shape)
-			fz_dump_blend(ctx, "/", state[0].shape);
+			fz_dump_blend(ctx, "/S=", state[0].shape);
+		if (state[0].group_alpha)
+			fz_dump_blend(ctx, "/GA=", state[0].group_alpha);
 		fz_dump_blend(ctx, " with ", state[1].mask);
 #endif
 		fz_paint_pixmap_with_mask(state[0].dest, state[1].dest, state[1].mask);
@@ -1915,6 +2025,11 @@ fz_draw_pop_clip(fz_context *ctx, fz_device *devp)
 			fz_paint_pixmap_with_mask(state[0].shape, state[1].shape, state[1].mask);
 			fz_drop_pixmap(ctx, state[1].shape);
 		}
+		if (state[0].group_alpha != state[1].group_alpha)
+		{
+			fz_paint_pixmap_with_mask(state[0].group_alpha, state[1].group_alpha, state[1].mask);
+			fz_drop_pixmap(ctx, state[1].group_alpha);
+		}
 		/* The following tests should not be required, but just occasionally
 		 * errors can cause the stack to get out of sync, and this might save
 		 * our bacon. */
@@ -1925,7 +2040,9 @@ fz_draw_pop_clip(fz_context *ctx, fz_device *devp)
 #ifdef DUMP_GROUP_BLENDS
 		fz_dump_blend(ctx, " to get ", state[0].dest);
 		if (state[0].shape)
-			fz_dump_blend(ctx, "/", state[0].shape);
+			fz_dump_blend(ctx, "/S=", state[0].shape);
+		if (state[0].group_alpha)
+			fz_dump_blend(ctx, "/GA=", state[0].group_alpha);
 		printf("\n");
 #endif
 	}
@@ -1945,6 +2062,7 @@ fz_draw_begin_mask(fz_context *ctx, fz_device *devp, const fz_rect *rect, int lu
 	fz_irect bbox;
 	fz_draw_state *state = push_stack(ctx, dev);
 	fz_pixmap *shape = state->shape;
+	fz_pixmap *group_alpha = state->group_alpha;
 	fz_rect trect = *rect;
 	fz_colorspace *colorspace = NULL;
 
@@ -1984,6 +2102,10 @@ fz_draw_begin_mask(fz_context *ctx, fz_device *devp, const fz_rect *rect, int lu
 			 */
 			state[1].shape = shape = NULL;
 		}
+		if (state->group_alpha)
+		{
+			state[1].group_alpha = group_alpha = NULL;
+		}
 
 		if (luminosity)
 		{
@@ -1994,12 +2116,16 @@ fz_draw_begin_mask(fz_context *ctx, fz_device *devp, const fz_rect *rect, int lu
 			fz_clear_pixmap_with_value(ctx, dest, bc * 255);
 			if (shape)
 				fz_clear_pixmap_with_value(ctx, shape, 255);
+			if (group_alpha)
+				fz_clear_pixmap_with_value(ctx, group_alpha, 255);
 		}
 		else
 		{
 			fz_clear_pixmap(ctx, dest);
 			if (shape)
 				fz_clear_pixmap(ctx, shape);
+			if (group_alpha)
+				fz_clear_pixmap(ctx, group_alpha);
 		}
 
 #ifdef DUMP_GROUP_BLENDS
@@ -2033,7 +2159,9 @@ fz_draw_end_mask(fz_context *ctx, fz_device *devp)
 	dump_spaces(dev->top-1, "Mask -> Clip: ");
 	fz_dump_blend(ctx, "Mask ", state[1].dest);
 	if (state[1].shape)
-		fz_dump_blend(ctx, "/", state[1].shape);
+		fz_dump_blend(ctx, "/S=", state[1].shape);
+	if (state[1].group_alpha)
+		fz_dump_blend(ctx, "/GA=", state[1].group_alpha);
 #endif
 	fz_try(ctx)
 	{
@@ -2048,6 +2176,9 @@ fz_draw_end_mask(fz_context *ctx, fz_device *devp)
 		if (state[1].shape != state[0].shape)
 			fz_drop_pixmap(ctx, state[1].shape);
 		state[1].shape = NULL;
+		if (state[1].group_alpha != state[0].group_alpha)
+			fz_drop_pixmap(ctx, state[1].group_alpha);
+		state[1].group_alpha = NULL;
 
 #ifdef DUMP_GROUP_BLENDS
 		fz_dump_blend(ctx, "-> Clip ", temp);
@@ -2069,6 +2200,11 @@ fz_draw_end_mask(fz_context *ctx, fz_device *devp)
 			state[1].shape = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
 			fz_clear_pixmap(ctx, state[1].shape);
 		}
+		if (state[0].group_alpha)
+		{
+			state[1].group_alpha = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
+			fz_clear_pixmap(ctx, state[1].group_alpha);
+		}
 		state[1].scissor = bbox;
 	}
 	fz_catch(ctx)
@@ -2113,18 +2249,13 @@ fz_draw_begin_group(fz_context *ctx, fz_device *devp, const fz_rect *rect, fz_co
 		if (isolated)
 		{
 			fz_clear_pixmap(ctx, dest);
+			state[1].group_alpha = NULL;
 		}
 		else
 		{
 			fz_copy_pixmap_rect(ctx, dest, state[0].dest, &bbox, dev->default_cs);
-		}
-
-		if (isolated)
-			state[1].shape = NULL;
-		else
-		{
-			state[1].shape = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
-			fz_clear_pixmap(ctx, state[1].shape);
+			state[1].group_alpha = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
+			fz_clear_pixmap(ctx, state[1].group_alpha);
 		}
 
 		state[1].alpha = alpha;
@@ -2135,18 +2266,20 @@ fz_draw_begin_group(fz_context *ctx, fz_device *devp, const fz_rect *rect, fz_co
 			char atext[80];
 			char btext[80];
 			if (alpha != 1)
-				sprintf(atext, " (alpha=%g)", alpha);
+				sprintf(atext, " (alpha %g)", alpha);
 			else
 				atext[0] = 0;
 			if (blendmode != 0)
-				sprintf(btext, " (blendmode=%d)", blendmode);
+				sprintf(btext, " (blend %d)", blendmode);
 			else
 				btext[0] = 0;
 			sprintf(text, "Group begin%s%s%s%s: background is ", isolated ? " (isolated)" : "", knockout ? " (knockout)" : "", atext, btext);
 			fz_dump_blend(ctx, text, state[1].dest);
 		}
 		if (state[1].shape)
-			fz_dump_blend(ctx, "/", state[1].shape);
+			fz_dump_blend(ctx, "/S=", state[1].shape);
+		if (state[1].group_alpha)
+			fz_dump_blend(ctx, "/GA=", state[1].group_alpha);
 		printf("\n");
 #endif
 
@@ -2183,10 +2316,14 @@ fz_draw_end_group(fz_context *ctx, fz_device *devp)
 	dump_spaces(dev->top, "");
 	fz_dump_blend(ctx, "Group end: blending ", state[1].dest);
 	if (state[1].shape)
-		fz_dump_blend(ctx, "/", state[1].shape);
+		fz_dump_blend(ctx, "/S=", state[1].shape);
+	if (state[1].group_alpha)
+		fz_dump_blend(ctx, "/GA=", state[1].group_alpha);
 	fz_dump_blend(ctx, " onto ", state[0].dest);
 	if (state[0].shape)
-		fz_dump_blend(ctx, "/", state[0].shape);
+		fz_dump_blend(ctx, "/S=", state[0].shape);
+	if (state[0].group_alpha)
+		fz_dump_blend(ctx, "/GA=", state[0].group_alpha);
 	if (alpha != 1.0f)
 		printf(" (alpha %g)", alpha);
 	if (blendmode != 0)
@@ -2203,20 +2340,33 @@ fz_draw_end_group(fz_context *ctx, fz_device *devp)
 		state[1].dest = converted;
 	}
 
-	if ((blendmode == 0) && (state[0].shape == state[1].shape))
+	if ((blendmode == 0) && (state[0].shape == state[1].shape) && (state[0].group_alpha == state[1].group_alpha))
 		fz_paint_pixmap(state[0].dest, state[1].dest, alpha * 255);
 	else
-		fz_blend_pixmap(ctx, state[0].dest, state[1].dest, alpha * 255, blendmode, isolated, state[1].shape);
+		fz_blend_pixmap(ctx, state[0].dest, state[1].dest, alpha * 255, blendmode, isolated, state[1].group_alpha);
 
-	assert(state[0].shape == NULL || state[0].shape != state[1].shape);
-	if (state[0].shape && state[0].shape != state[1].shape)
+	if (state[0].shape != state[1].shape)
 	{
-		if (state[1].shape)
-			fz_paint_pixmap(state[0].shape, state[1].shape, alpha * 255);
+		if (state[0].shape)
+		{
+			if (state[1].shape)
+				fz_paint_pixmap(state[0].shape, state[1].shape, 255);
+			else
+				fz_paint_pixmap_alpha(state[0].shape, state[1].dest, 255);
+		}
+		fz_drop_pixmap(ctx, state[1].shape);
+	}
+	assert(state[0].group_alpha == NULL || state[0].group_alpha != state[1].group_alpha);
+	if (state[0].group_alpha && state[0].group_alpha != state[1].group_alpha)
+	{
+		/* The 'D' on page 7 of Altona_Technical_v20_x4.pdf goes wrong if this
+		 * is 255 * alpha, as an alpha effectively gets applied twice. */
+		if (state[1].group_alpha)
+			fz_paint_pixmap(state[0].group_alpha, state[1].group_alpha, 255);
 		else
-			fz_paint_pixmap_alpha(state[0].shape, state[1].dest, alpha * 255);
+			fz_paint_pixmap_alpha(state[0].group_alpha, state[1].dest, 255);
 	}
-	fz_drop_pixmap(ctx, state[1].shape);
+	fz_drop_pixmap(ctx, state[1].group_alpha);
 	/* The following test should not be required, but just occasionally
 	 * errors can cause the stack to get out of sync, and this might save
 	 * our bacon. */
@@ -2226,7 +2376,9 @@ fz_draw_end_group(fz_context *ctx, fz_device *devp)
 #ifdef DUMP_GROUP_BLENDS
 	fz_dump_blend(ctx, " to get ", state[0].dest);
 	if (state[0].shape)
-		fz_dump_blend(ctx, "/", state[0].shape);
+		fz_dump_blend(ctx, "/S=", state[0].shape);
+	if (state[0].group_alpha)
+		fz_dump_blend(ctx, "/GA=", state[0].group_alpha);
 	printf("\n");
 #endif
 
@@ -2247,6 +2399,7 @@ typedef struct
 	fz_storable storable;
 	fz_pixmap *dest;
 	fz_pixmap *shape;
+	fz_pixmap *group_alpha;
 } tile_record;
 
 static int
@@ -2318,6 +2471,7 @@ fz_drop_tile_record_imp(fz_context *ctx, fz_storable *storable)
 	tile_record *tr = (tile_record *)storable;
 	fz_drop_pixmap(ctx, tr->dest);
 	fz_drop_pixmap(ctx, tr->shape);
+	fz_drop_pixmap(ctx, tr->group_alpha);
 	fz_free(ctx, tr);
 }
 
@@ -2328,12 +2482,13 @@ fz_drop_tile_record(fz_context *ctx, tile_record *tile)
 }
 
 static tile_record *
-fz_new_tile_record(fz_context *ctx, fz_pixmap *dest, fz_pixmap *shape)
+fz_new_tile_record(fz_context *ctx, fz_pixmap *dest, fz_pixmap *shape, fz_pixmap *group_alpha)
 {
 	tile_record *tile = fz_malloc_struct(ctx, tile_record);
 	FZ_INIT_STORABLE(tile, 1, fz_drop_tile_record_imp);
 	tile->dest = fz_keep_pixmap(ctx, dest);
 	tile->shape = fz_keep_pixmap(ctx, shape);
+	tile->group_alpha = fz_keep_pixmap(ctx, group_alpha);
 	return tile;
 }
 
@@ -2342,7 +2497,7 @@ fz_tile_size(fz_context *ctx, tile_record *tile)
 {
 	if (!tile)
 		return 0;
-	return sizeof(*tile) + fz_pixmap_size(ctx, tile->dest) + fz_pixmap_size(ctx, tile->shape);
+	return sizeof(*tile) + fz_pixmap_size(ctx, tile->dest) + fz_pixmap_size(ctx, tile->shape) + fz_pixmap_size(ctx, tile->group_alpha);
 }
 
 static int
@@ -2351,7 +2506,7 @@ fz_draw_begin_tile(fz_context *ctx, fz_device *devp, const fz_rect *area, const
 	fz_draw_device *dev = (fz_draw_device*)devp;
 	fz_matrix ctm = concat(in_ctm, &dev->transform);
 	fz_pixmap *dest = NULL;
-	fz_pixmap *shape;
+	fz_pixmap *shape, *group_alpha;
 	fz_irect bbox;
 	fz_draw_state *state = &dev->stack[dev->top];
 	fz_colorspace *model = state->dest->colorspace;
@@ -2394,6 +2549,7 @@ fz_draw_begin_tile(fz_context *ctx, fz_device *devp, const fz_rect *area, const
 		{
 			state[1].dest = fz_keep_pixmap(ctx, tile->dest);
 			state[1].shape = fz_keep_pixmap(ctx, tile->shape);
+			state[1].group_alpha = fz_keep_pixmap(ctx, tile->group_alpha);
 			state[1].blendmode |= FZ_BLEND_ISOLATED;
 			state[1].xstep = xstep;
 			state[1].ystep = ystep;
@@ -2422,6 +2578,12 @@ fz_draw_begin_tile(fz_context *ctx, fz_device *devp, const fz_rect *area, const
 			state[1].shape = shape = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
 			fz_clear_pixmap(ctx, shape);
 		}
+		group_alpha = state[0].group_alpha;
+		if (group_alpha)
+		{
+			state[1].group_alpha = group_alpha = fz_new_pixmap_with_bbox(ctx, NULL, &bbox, NULL, 1);
+			fz_clear_pixmap(ctx, group_alpha);
+		}
 		state[1].blendmode |= FZ_BLEND_ISOLATED;
 		state[1].xstep = xstep;
 		state[1].ystep = ystep;
@@ -2448,7 +2610,7 @@ fz_draw_end_tile(fz_context *ctx, fz_device *devp)
 {
 	fz_draw_device *dev = (fz_draw_device*)devp;
 	float xstep, ystep;
-	fz_matrix ttm, ctm, shapectm;
+	fz_matrix ttm, ctm, shapectm, gactm;
 	fz_irect area, scissor, tile_bbox;
 	fz_rect scissor_tmp, tile_tmp;
 	int x0, y0, x1, y1, x, y, extra_x, extra_y;
@@ -2505,15 +2667,25 @@ fz_draw_end_tile(fz_context *ctx, fz_device *devp)
 		shapectm.e = state[1].shape->x;
 		shapectm.f = state[1].shape->y;
 	}
+	if (state[1].group_alpha)
+	{
+		gactm = ctm;
+		gactm.e = state[1].group_alpha->x;
+		gactm.f = state[1].group_alpha->y;
+	}
 
 #ifdef DUMP_GROUP_BLENDS
 	dump_spaces(dev->top, "");
 	fz_dump_blend(ctx, "Tiling ", state[1].dest);
 	if (state[1].shape)
-		fz_dump_blend(ctx, "/", state[1].shape);
+		fz_dump_blend(ctx, "/S=", state[1].shape);
+	if (state[1].group_alpha)
+		fz_dump_blend(ctx, "/GA=", state[1].group_alpha);
 	fz_dump_blend(ctx, " onto ", state[0].dest);
 	if (state[0].shape)
-		fz_dump_blend(ctx, "/", state[0].shape);
+		fz_dump_blend(ctx, "/S=", state[0].shape);
+	if (state[0].group_alpha)
+		fz_dump_blend(ctx, "/GA=", state[0].group_alpha);
 #endif
 
 	for (y = y0; y < y1; y++)
@@ -2538,6 +2710,14 @@ fz_draw_end_tile(fz_context *ctx, fz_device *devp)
 				state[1].shape->y = ttm.f;
 				fz_paint_pixmap_with_bbox(state[0].shape, state[1].shape, 255, state[0].scissor);
 			}
+			if (state[1].group_alpha)
+			{
+				ttm = gactm;
+				fz_pre_translate(&ttm, x * xstep, y * ystep);
+				state[1].group_alpha->x = ttm.e;
+				state[1].group_alpha->y = ttm.f;
+				fz_paint_pixmap_with_bbox(state[0].group_alpha, state[1].group_alpha, 255, state[0].scissor);
+			}
 		}
 	}
 
@@ -2548,6 +2728,11 @@ fz_draw_end_tile(fz_context *ctx, fz_device *devp)
 		state[1].shape->x = shapectm.e;
 		state[1].shape->y = shapectm.f;
 	}
+	if (state[1].group_alpha)
+	{
+		state[1].group_alpha->x = gactm.e;
+		state[1].group_alpha->y = gactm.f;
+	}
 
 	/* Now we try to cache the tiles. Any failure here will just result in us not caching. */
 	if (state[1].encache && state[1].id != 0)
@@ -2560,7 +2745,7 @@ fz_draw_end_tile(fz_context *ctx, fz_device *devp)
 		{
 			tile_record *existing_tile;
 
-			tile = fz_new_tile_record(ctx, state[1].dest, state[1].shape);
+			tile = fz_new_tile_record(ctx, state[1].dest, state[1].shape, state[1].group_alpha);
 
 			key = fz_malloc_struct(ctx, tile_key);
 			key->refs = 1;
@@ -2598,10 +2783,14 @@ fz_draw_end_tile(fz_context *ctx, fz_device *devp)
 		fz_drop_pixmap(ctx, state[1].dest);
 	if (state[0].shape != state[1].shape)
 		fz_drop_pixmap(ctx, state[1].shape);
+	if (state[0].group_alpha != state[1].group_alpha)
+		fz_drop_pixmap(ctx, state[1].group_alpha);
 #ifdef DUMP_GROUP_BLENDS
 	fz_dump_blend(ctx, " to get ", state[0].dest);
 	if (state[0].shape)
-		fz_dump_blend(ctx, "/", state[0].shape);
+		fz_dump_blend(ctx, "/S=", state[0].shape);
+	if (state[0].group_alpha)
+		fz_dump_blend(ctx, "/GA=", state[0].group_alpha);
 	printf("\n");
 #endif
 
@@ -2643,6 +2832,8 @@ fz_draw_close_device(fz_context *ctx, fz_device *devp)
 			fz_drop_pixmap(ctx, state[1].dest);
 		if (state[1].shape != state[0].shape)
 			fz_drop_pixmap(ctx, state[1].shape);
+		if (state[1].group_alpha != state[0].group_alpha)
+			fz_drop_pixmap(ctx, state[1].group_alpha);
 	}
 
 	if (dev->resolve_spots && dev->top)
@@ -2652,6 +2843,7 @@ fz_draw_close_device(fz_context *ctx, fz_device *devp)
 		fz_drop_pixmap(ctx, state[1].dest);
 		assert(state[1].mask == NULL);
 		assert(state[1].shape == NULL);
+		assert(state[1].group_alpha == NULL);
 	}
 }
 
@@ -2676,6 +2868,8 @@ fz_draw_drop_device(fz_context *ctx, fz_device *devp)
 			fz_drop_pixmap(ctx, state[1].dest);
 		if (state[1].shape != state[0].shape)
 			fz_drop_pixmap(ctx, state[1].shape);
+		if (state[1].group_alpha != state[0].group_alpha)
+			fz_drop_pixmap(ctx, state[1].group_alpha);
 	}
 
 	/* We never free the dest/mask/shape at level 0, as:
@@ -2734,6 +2928,7 @@ new_draw_device(fz_context *ctx, const fz_matrix *transform, fz_pixmap *dest, co
 	dev->stack_cap = STACK_SIZE;
 	dev->stack[0].dest = dest;
 	dev->stack[0].shape = NULL;
+	dev->stack[0].group_alpha = NULL;
 	dev->stack[0].mask = NULL;
 	dev->stack[0].blendmode = 0;
 	dev->stack[0].scissor.x0 = dest->x;
diff --git a/source/fitz/draw-imp.h b/source/fitz/draw-imp.h
index d2d50d2e..e855b40a 100644
--- a/source/fitz/draw-imp.h
+++ b/source/fitz/draw-imp.h
@@ -454,8 +454,8 @@ fz_solid_color_painter_t *fz_get_solid_color_painter(int n, const unsigned char
 fz_span_painter_t *fz_get_span_painter(int da, int sa, int n, int alpha, const fz_overprint * restrict eop);
 fz_span_color_painter_t *fz_get_span_color_painter(int n, int da, const unsigned char * restrict color, const fz_overprint * restrict eop);
 
-void fz_paint_image(fz_pixmap * restrict dst, const fz_irect * restrict scissor, fz_pixmap * restrict shape, const fz_pixmap * restrict img, const fz_matrix * restrict ctm, int alpha, int lerp_allowed, int gridfit_as_tiled, const fz_overprint * restrict eop);
-void fz_paint_image_with_color(fz_pixmap * restrict dst, const fz_irect * restrict scissor, fz_pixmap *restrict shape, const fz_pixmap * restrict img, const fz_matrix * restrict ctm, const unsigned char * restrict colorbv, int lerp_allowed, int gridfit_as_tiled, const fz_overprint * restrict eop);
+void fz_paint_image(fz_pixmap * restrict dst, const fz_irect * restrict scissor, fz_pixmap * restrict shape, fz_pixmap * restrict group_alpha, const fz_pixmap * restrict img, const fz_matrix * restrict ctm, int alpha, int lerp_allowed, int gridfit_as_tiled, const fz_overprint * restrict eop);
+void fz_paint_image_with_color(fz_pixmap * restrict dst, const fz_irect * restrict scissor, fz_pixmap *restrict shape, fz_pixmap * restrict group_alpha, const fz_pixmap * restrict img, const fz_matrix * restrict ctm, const unsigned char * restrict colorbv, int lerp_allowed, int gridfit_as_tiled, const fz_overprint * restrict eop);
 
 void fz_paint_pixmap(fz_pixmap * restrict dst, const fz_pixmap * restrict src, int alpha);
 void fz_paint_pixmap_alpha(fz_pixmap * restrict dst, const fz_pixmap * restrict src, int alpha);