Outline of new features

Signed-off-by: Andrew Opalach <andrew@akon.city>

1 files changed, 191 insertions, 0 deletions

diff --git a/Shaders/volume_upsample2.diff b/Shaders/volume_upsample2.diff
new file mode 100644
index 0000000..0887f10
--- /dev/null
+++ b/Shaders/volume_upsample2.diff
@@ -0,0 +1,191 @@
+--- a/volume_upsample2.asm	2026-03-16 15:19:50.820204181 -0400
++++ b/volume_upsample2.asm	2026-03-16 15:19:31.707007465 -0400
+@@ -254,103 +182,92 @@
+ else
+   mov r0.w, l(0)
+ endif
+-if_nz r0.w
+-  utof r1.xy, r2.xyxx
+-  mul r1.xy, r1.xyxx, cb3[2].zwzz
+-  mul r1.xy, r1.xyxx, cb1[4].zzzz
+-  mad r1.zw, -cb1[1].zzzw, l(0.000000, 0.000000, 0.500000, 0.500000), cb1[4].zzzz
+-  max r1.xy, r1.xyxx, l(0.000000, 0.000000, 0.000000, 0.000000)
+-  min r1.xy, r1.zwzz, r1.xyxx
+-  sample_l_indexable(texture2d)(float,float,float,float) r3.xyz, r1.xyxx, t1.xyzw, s0, l(0.000000)
+-  sample_l_indexable(texture2d)(float,float,float,float) r1.xyz, r1.xyxx, t2.xyzw, s0, l(0.000000)
++if_nz r0.z
++  iadd r0.xy, -r0.xyxx, r2.yxyy
++  mov r4.xy, vThreadID.xyxx
++  mov r4.zw, l(0,0,0,0)
++  ld_indexable(texture2d)(float,float,float,float) r0.z, r4.xyzw, t0.yzxw
++  not r4.xy, r0.xyxx
++  imad r0.y, r4.x, l(-9), r0.y
++  imad r0.w, r4.x, l(-9), -r4.y
++  ishl r0.w, r0.w, l(8)
++  or r0.w, r0.w, r0.y
++  imad r0.x, r0.x, l(9), -r4.y
++  ishl r0.xy, r0.xyxx, l(16, 24, 0, 0)
++  or r0.x, r0.x, r0.w
++  or r0.x, r0.y, r0.x
++  bfi r0.y, l(1), l(1), vThreadIDInGroup.y, l(0)
++  bfi r0.y, l(1), l(0), vThreadIDInGroup.x, r0.y
++  and r0.w, r0.x, l(255)
++  ld_structured r1.w, r0.w, l(0), g0.xxxx
++  add r1.w, r0.z, -r1.w
++  add r1.w, |r1.w|, l(1.192092896E-07)
++  iadd r3.w, -r1.w, l(0x7ef312ac)
++  mad r4.x, -r3.w, r1.w, l(2.000000)
++  mul r3.w, r3.w, r4.x
++  mad r1.w, -r3.w, r1.w, l(2.000000)
++  mul r1.w, r1.w, r3.w
++  mul r3.w, r1.w, icb[r0.y + 0].x
++  ld_structured r4.xyz, r0.w, l(0), g1.xyzx
++  ld_structured r5.xyz, r0.w, l(0), g2.xyzx
++  mad r0.w, r1.w, icb[r0.y + 0].x, l(1.192092896E-07)
++  ubfe r6.xy, l(8, 8, 0, 0), l(8, 16, 0, 0), r0.xxxx
++  ld_structured r1.w, r6.x, l(0), g0.xxxx
++  add r1.w, r0.z, -r1.w
++  add r1.w, |r1.w|, l(1.192092896E-07)
++  iadd r4.w, -r1.w, l(0x7ef312ac)
++  mad r5.w, -r4.w, r1.w, l(2.000000)
++  mul r4.w, r4.w, r5.w
++  mad r1.w, -r4.w, r1.w, l(2.000000)
++  mul r1.w, r1.w, r4.w
++  mul r4.w, r1.w, icb[r0.y + 0].y
++  ld_structured r7.xyz, r6.x, l(0), g1.xyzx
++  mul r7.xyz, r4.wwww, r7.xyzx
++  mad r4.xyz, r4.xyzx, r3.wwww, r7.xyzx
++  ld_structured r6.xzw, r6.x, l(0), g2.xxyz
++  mul r6.xzw, r4.wwww, r6.xxzw
++  mad r5.xyz, r5.xyzx, r3.wwww, r6.xzwx
++  mad r0.w, r1.w, icb[r0.y + 0].y, r0.w
++  ld_structured r1.w, r6.y, l(0), g0.xxxx
++  add r1.w, r0.z, -r1.w
++  add r1.w, |r1.w|, l(1.192092896E-07)
++  iadd r3.w, -r1.w, l(0x7ef312ac)
++  mad r4.w, -r3.w, r1.w, l(2.000000)
++  mul r3.w, r3.w, r4.w
++  mad r1.w, -r3.w, r1.w, l(2.000000)
++  mul r1.w, r1.w, r3.w
++  mul r3.w, r1.w, icb[r0.y + 0].z
++  ld_structured r6.xzw, r6.y, l(0), g1.xxyz
++  mad r4.xyz, r6.xzwx, r3.wwww, r4.xyzx
++  ld_structured r6.xyz, r6.y, l(0), g2.xyzx
++  mad r5.xyz, r6.xyzx, r3.wwww, r5.xyzx
++  mad r0.w, r1.w, icb[r0.y + 0].z, r0.w
++  ushr r0.x, r0.x, l(24)
++  ld_structured r1.w, r0.x, l(0), g0.xxxx
++  add r0.z, r0.z, -r1.w
++  add r0.z, |r0.z|, l(1.192092896E-07)
++  iadd r1.w, -r0.z, l(0x7ef312ac)
++  mad r3.w, -r1.w, r0.z, l(2.000000)
++  mul r1.w, r1.w, r3.w
++  mad r0.z, -r1.w, r0.z, l(2.000000)
++  mul r0.z, r0.z, r1.w
++  mul r1.w, r0.z, icb[r0.y + 0].w
++  ld_structured r6.xyz, r0.x, l(0), g1.xyzx
++  mad r4.xyz, r6.xyzx, r1.wwww, r4.xyzx
++  ld_structured r6.xyz, r0.x, l(0), g2.xyzx
++  mad r5.xyz, r6.xyzx, r1.wwww, r5.xyzx
++  mad r0.x, r0.z, icb[r0.y + 0].w, r0.w
++  iadd r0.y, -r0.x, l(0x7ef312ac)
++  mad r0.z, -r0.y, r0.x, l(2.000000)
++  mul r0.y, r0.z, r0.y
++  mad r0.x, -r0.y, r0.x, l(2.000000)
++  mul r0.x, r0.x, r0.y
++  mul r3.xyz, r0.xxxx, r4.xyzx
++  mul r1.xyz, r0.xxxx, r5.xyzx
+ else
+-  if_nz r0.z
+-    iadd r0.xy, -r0.xyxx, r2.yxyy
+-    mov r4.xy, vThreadID.xyxx
+-    mov r4.zw, l(0,0,0,0)
+-    ld_indexable(texture2d)(float,float,float,float) r0.z, r4.xyzw, t0.yzxw
+-    not r4.xy, r0.xyxx
+-    imad r0.y, r4.x, l(-9), r0.y
+-    imad r0.w, r4.x, l(-9), -r4.y
+-    ishl r0.w, r0.w, l(8)
+-    or r0.w, r0.w, r0.y
+-    imad r0.x, r0.x, l(9), -r4.y
+-    ishl r0.xy, r0.xyxx, l(16, 24, 0, 0)
+-    or r0.x, r0.x, r0.w
+-    or r0.x, r0.y, r0.x
+-    bfi r0.y, l(1), l(1), vThreadIDInGroup.y, l(0)
+-    bfi r0.y, l(1), l(0), vThreadIDInGroup.x, r0.y
+-    and r0.w, r0.x, l(255)
+-    ld_structured r1.w, r0.w, l(0), g0.xxxx
+-    add r1.w, r0.z, -r1.w
+-    add r1.w, |r1.w|, l(1.192092896E-07)
+-    iadd r3.w, -r1.w, l(0x7ef312ac)
+-    mad r4.x, -r3.w, r1.w, l(2.000000)
+-    mul r3.w, r3.w, r4.x
+-    mad r1.w, -r3.w, r1.w, l(2.000000)
+-    mul r1.w, r1.w, r3.w
+-    mul r3.w, r1.w, icb[r0.y + 0].x
+-    ld_structured r4.xyz, r0.w, l(0), g1.xyzx
+-    ld_structured r5.xyz, r0.w, l(0), g2.xyzx
+-    mad r0.w, r1.w, icb[r0.y + 0].x, l(1.192092896E-07)
+-    ubfe r6.xy, l(8, 8, 0, 0), l(8, 16, 0, 0), r0.xxxx
+-    ld_structured r1.w, r6.x, l(0), g0.xxxx
+-    add r1.w, r0.z, -r1.w
+-    add r1.w, |r1.w|, l(1.192092896E-07)
+-    iadd r4.w, -r1.w, l(0x7ef312ac)
+-    mad r5.w, -r4.w, r1.w, l(2.000000)
+-    mul r4.w, r4.w, r5.w
+-    mad r1.w, -r4.w, r1.w, l(2.000000)
+-    mul r1.w, r1.w, r4.w
+-    mul r4.w, r1.w, icb[r0.y + 0].y
+-    ld_structured r7.xyz, r6.x, l(0), g1.xyzx
+-    mul r7.xyz, r4.wwww, r7.xyzx
+-    mad r4.xyz, r4.xyzx, r3.wwww, r7.xyzx
+-    ld_structured r6.xzw, r6.x, l(0), g2.xxyz
+-    mul r6.xzw, r4.wwww, r6.xxzw
+-    mad r5.xyz, r5.xyzx, r3.wwww, r6.xzwx
+-    mad r0.w, r1.w, icb[r0.y + 0].y, r0.w
+-    ld_structured r1.w, r6.y, l(0), g0.xxxx
+-    add r1.w, r0.z, -r1.w
+-    add r1.w, |r1.w|, l(1.192092896E-07)
+-    iadd r3.w, -r1.w, l(0x7ef312ac)
+-    mad r4.w, -r3.w, r1.w, l(2.000000)
+-    mul r3.w, r3.w, r4.w
+-    mad r1.w, -r3.w, r1.w, l(2.000000)
+-    mul r1.w, r1.w, r3.w
+-    mul r3.w, r1.w, icb[r0.y + 0].z
+-    ld_structured r6.xzw, r6.y, l(0), g1.xxyz
+-    mad r4.xyz, r6.xzwx, r3.wwww, r4.xyzx
+-    ld_structured r6.xyz, r6.y, l(0), g2.xyzx
+-    mad r5.xyz, r6.xyzx, r3.wwww, r5.xyzx
+-    mad r0.w, r1.w, icb[r0.y + 0].z, r0.w
+-    ushr r0.x, r0.x, l(24)
+-    ld_structured r1.w, r0.x, l(0), g0.xxxx
+-    add r0.z, r0.z, -r1.w
+-    add r0.z, |r0.z|, l(1.192092896E-07)
+-    iadd r1.w, -r0.z, l(0x7ef312ac)
+-    mad r3.w, -r1.w, r0.z, l(2.000000)
+-    mul r1.w, r1.w, r3.w
+-    mad r0.z, -r1.w, r0.z, l(2.000000)
+-    mul r0.z, r0.z, r1.w
+-    mul r1.w, r0.z, icb[r0.y + 0].w
+-    ld_structured r6.xyz, r0.x, l(0), g1.xyzx
+-    mad r4.xyz, r6.xyzx, r1.wwww, r4.xyzx
+-    ld_structured r6.xyz, r0.x, l(0), g2.xyzx
+-    mad r5.xyz, r6.xyzx, r1.wwww, r5.xyzx
+-    mad r0.x, r0.z, icb[r0.y + 0].w, r0.w
+-    iadd r0.y, -r0.x, l(0x7ef312ac)
+-    mad r0.z, -r0.y, r0.x, l(2.000000)
+-    mul r0.y, r0.z, r0.y
+-    mad r0.x, -r0.y, r0.x, l(2.000000)
+-    mul r0.x, r0.x, r0.y
+-    mul r3.xyz, r0.xxxx, r4.xyzx
+-    mul r1.xyz, r0.xxxx, r5.xyzx
+-  else
+-    mov r2.zw, l(0,0,0,0)
+-    ld_indexable(texture2d)(float,float,float,float) r3.xyz, r2.xyww, t1.xyzw
+-    ld_indexable(texture2d)(float,float,float,float) r1.xyz, r2.xyzw, t2.xyzw
+-  endif
++  mov r2.zw, l(0,0,0,0)
++  ld_indexable(texture2d)(float,float,float,float) r3.xyz, r2.xyww, t1.xyzw
++  ld_indexable(texture2d)(float,float,float,float) r1.xyz, r2.xyzw, t2.xyzw
+ endif
+ store_uav_typed u1.xyzw, vThreadID.xyyy, r1.xyzz
+ store_uav_typed u0.xyzw, vThreadID.xyyy, r3.xyzz