pp: Add Jimenez' MLAA

Signed-off-by: Lauri Kasanen <cand@gmx.com>
Signed-off-by: Brian Paul <brianp@vmware.com>

2 files changed, 646 insertions, 0 deletions

diff --git a/src/gallium/auxiliary/postprocess/pp_mlaa.c b/src/gallium/auxiliary/postprocess/pp_mlaa.c
new file mode 100644
index 00000000000..476502fca93
--- /dev/null
+++ b/src/gallium/auxiliary/postprocess/pp_mlaa.c
@@ -0,0 +1,304 @@
+/**
+ * Copyright (C) 2010 Jorge Jimenez (jorge@iryoku.com)
+ * Copyright (C) 2010 Belen Masia (bmasia@unizar.es)
+ * Copyright (C) 2010 Jose I. Echevarria (joseignacioechevarria@gmail.com)
+ * Copyright (C) 2010 Fernando Navarro (fernandn@microsoft.com)
+ * Copyright (C) 2010 Diego Gutierrez (diegog@unizar.es)
+ * Copyright (C) 2011 Lauri Kasanen (cand@gmx.com)
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ *    1. Redistributions of source code must retain the above copyright notice,
+ *       this list of conditions and the following disclaimer.
+ *
+ *    2. Redistributions in binary form must reproduce the following statement:
+ *
+ *       "Uses Jimenez's MLAA. Copyright (C) 2010 by Jorge Jimenez, Belen Masia,
+ *        Jose I. Echevarria, Fernando Navarro and Diego Gutierrez."
+ *
+ *       Only for use in the Mesa project, this point 2 is filled by naming the
+ *       technique Jimenez's MLAA in the Mesa config options.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS
+ * IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
+ * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL COPYRIGHT HOLDERS OR CONTRIBUTORS
+ * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ *
+ * The views and conclusions contained in the software and documentation are
+ * those of the authors and should not be interpreted as representing official
+ * policies, either expressed or implied, of the copyright holders.
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include "postprocess/postprocess.h"
+#include "postprocess/pp_mlaa.h"
+#include "postprocess/pp_filters.h"
+#include "util/u_blit.h"
+#include "util/u_box.h"
+#include "util/u_sampler.h"
+#include "util/u_inlines.h"
+#include "pipe/p_screen.h"
+
+#define IMM_SPACE 80
+
+static float constants[] = { 1, 1, 0, 0 };
+static unsigned int dimensions[2] = { 0, 0 };
+
+static struct pipe_resource *constbuf, *areamaptex;
+
+/** Upload the constants. */
+static void
+up_consts(struct pipe_context *pipe)
+{
+   struct pipe_box box;
+
+   u_box_2d(0, 0, sizeof(constants), 1, &box);
+   pipe->transfer_inline_write(pipe, constbuf, 0, PIPE_TRANSFER_WRITE,
+                               &box, constants, sizeof(constants),
+                               sizeof(constants));
+}
+
+/** Run function of the MLAA filter. */
+static void
+pp_jimenezmlaa_run(struct pp_queue_t *ppq, struct pipe_resource *in,
+                   struct pipe_resource *out, unsigned int n, bool iscolor)
+{
+
+   struct program *p = ppq->p;
+
+   struct pipe_depth_stencil_alpha_state mstencil;
+   struct pipe_sampler_view v_tmp, *arr[3];
+
+   unsigned int w = p->framebuffer.width;
+   unsigned int h = p->framebuffer.height;
+
+   const struct pipe_stencil_ref ref = { {1} };
+   memset(&mstencil, 0, sizeof(mstencil));
+   cso_set_stencil_ref(p->cso, &ref);
+
+   /* Init the pixel size constant */
+   if (dimensions[0] != p->framebuffer.width ||
+       dimensions[1] != p->framebuffer.height) {
+      constants[0] = 1.0 / p->framebuffer.width;
+      constants[1] = 1.0 / p->framebuffer.height;
+
+      up_consts(p->pipe);
+      dimensions[0] = p->framebuffer.width;
+      dimensions[1] = p->framebuffer.height;
+   }
+
+   p->pipe->set_constant_buffer(p->pipe, PIPE_SHADER_VERTEX, 0, constbuf);
+   p->pipe->set_constant_buffer(p->pipe, PIPE_SHADER_FRAGMENT, 0, constbuf);
+
+   mstencil.stencil[0].enabled = 1;
+   mstencil.stencil[0].valuemask = mstencil.stencil[0].writemask = ~0;
+   mstencil.stencil[0].func = PIPE_FUNC_ALWAYS;
+   mstencil.stencil[0].fail_op = PIPE_STENCIL_OP_KEEP;
+   mstencil.stencil[0].zfail_op = PIPE_STENCIL_OP_KEEP;
+   mstencil.stencil[0].zpass_op = PIPE_STENCIL_OP_REPLACE;
+
+   p->framebuffer.zsbuf = ppq->stencils;
+
+   /* First pass: depth edge detection */
+   if (iscolor)
+      pp_filter_setup_in(p, in);
+   else
+      pp_filter_setup_in(p, ppq->depth);
+
+   pp_filter_setup_out(p, ppq->inner_tmp[0]);
+
+   pp_filter_set_fb(p);
+   pp_filter_misc_state(p);
+   cso_set_depth_stencil_alpha(p->cso, &mstencil);
+   p->pipe->clear(p->pipe, PIPE_CLEAR_STENCIL | PIPE_CLEAR_COLOR,
+                  p->clear_color, 0, 0);
+
+   cso_single_sampler(p->cso, 0, &p->sampler_point);
+   cso_single_sampler_done(p->cso);
+   cso_set_fragment_sampler_views(p->cso, 1, &p->view);
+
+   cso_set_vertex_shader_handle(p->cso, ppq->shaders[n][1]);    /* offsetvs */
+   cso_set_fragment_shader_handle(p->cso, ppq->shaders[n][2]);
+
+   pp_filter_draw(p);
+   pp_filter_end_pass(p);
+
+
+   /* Second pass: blend weights */
+   /* Sampler order: areamap, edgesmap, edgesmapL (reversed, thx compiler) */
+   mstencil.stencil[0].func = PIPE_FUNC_EQUAL;
+   mstencil.stencil[0].zpass_op = PIPE_STENCIL_OP_KEEP;
+   cso_set_depth_stencil_alpha(p->cso, &mstencil);
+
+   pp_filter_setup_in(p, areamaptex);
+   pp_filter_setup_out(p, ppq->inner_tmp[1]);
+
+   u_sampler_view_default_template(&v_tmp, ppq->inner_tmp[0],
+                                   ppq->inner_tmp[0]->format);
+   arr[1] = arr[2] = p->pipe->create_sampler_view(p->pipe,
+                                                  ppq->inner_tmp[0], &v_tmp);
+
+   pp_filter_set_clear_fb(p);
+
+   cso_single_sampler(p->cso, 0, &p->sampler_point);
+   cso_single_sampler(p->cso, 1, &p->sampler_point);
+   cso_single_sampler(p->cso, 2, &p->sampler);
+   cso_single_sampler_done(p->cso);
+
+   arr[0] = p->view;
+   cso_set_fragment_sampler_views(p->cso, 3, arr);
+
+   cso_set_vertex_shader_handle(p->cso, ppq->shaders[n][0]);    /* passvs */
+   cso_set_fragment_shader_handle(p->cso, ppq->shaders[n][3]);
+
+   pp_filter_draw(p);
+   pp_filter_end_pass(p);
+   pipe_sampler_view_reference(&arr[1], NULL);
+
+
+   /* Third pass: smoothed edges */
+   /* Sampler order: colormap, blendmap (wtf compiler) */
+   pp_filter_setup_in(p, ppq->inner_tmp[1]);
+   pp_filter_setup_out(p, out);
+
+   pp_filter_set_fb(p);
+
+   /* Blit the input to the output */
+   util_blit_pixels(p->blitctx, in, 0, 0, 0,
+                    w, h, 0, p->framebuffer.cbufs[0],
+                    0, 0, w, h, 0, PIPE_TEX_MIPFILTER_NEAREST);
+
+   u_sampler_view_default_template(&v_tmp, in, in->format);
+   arr[0] = p->pipe->create_sampler_view(p->pipe, in, &v_tmp);
+
+   cso_single_sampler(p->cso, 0, &p->sampler_point);
+   cso_single_sampler(p->cso, 1, &p->sampler_point);
+   cso_single_sampler_done(p->cso);
+
+   arr[1] = p->view;
+   cso_set_fragment_sampler_views(p->cso, 2, arr);
+
+   cso_set_vertex_shader_handle(p->cso, ppq->shaders[n][1]);    /* offsetvs */
+   cso_set_fragment_shader_handle(p->cso, ppq->shaders[n][4]);
+
+   p->blend.rt[0].blend_enable = 1;
+   cso_set_blend(p->cso, &p->blend);
+
+   pp_filter_draw(p);
+   pp_filter_end_pass(p);
+   pipe_sampler_view_reference(&arr[0], NULL);
+
+   p->blend.rt[0].blend_enable = 0;
+   p->framebuffer.zsbuf = NULL;
+}
+
+/** The init function of the MLAA filter. */
+static void
+pp_jimenezmlaa_init_run(struct pp_queue_t *ppq, unsigned int n,
+                        unsigned int val, bool iscolor)
+{
+
+   struct pipe_box box;
+   struct pipe_resource res;
+
+   char *tmp_text = calloc(sizeof(blend2fs_1) + sizeof(blend2fs_2) +
+                           IMM_SPACE, sizeof(char));
+
+   constbuf = pipe_buffer_create(ppq->p->screen, PIPE_BIND_CONSTANT_BUFFER,
+                                 PIPE_USAGE_STATIC, sizeof(constants));
+   if (!constbuf) {
+      pp_debug("Failed to allocate constant buffer\n");
+      return;
+   }
+
+
+   pp_debug("mlaa: using %u max search steps\n", val);
+
+   if (!tmp_text) {
+      pp_debug("Failed to allocate shader space\n");
+      return;
+   }
+   sprintf(tmp_text, "%s"
+           "IMM FLT32 {    %.8f,     0.0000,     0.0000,     0.0000}\n"
+           "%s\n", blend2fs_1, (float) val, blend2fs_2);
+
+   memset(&res, 0, sizeof(res));
+
+   res.target = PIPE_TEXTURE_2D;
+   res.format = PIPE_FORMAT_R8G8_UNORM;
+   res.width0 = res.height0 = 165;
+   res.bind = PIPE_BIND_SAMPLER_VIEW;
+   res.usage = PIPE_USAGE_STATIC;
+   res.depth0 = res.array_size = res.nr_samples = 1;
+
+   if (!ppq->p->screen->is_format_supported(ppq->p->screen, res.format,
+                                            res.target, 1, res.bind))
+      pp_debug("Areamap format not supported\n");
+
+   areamaptex = ppq->p->screen->resource_create(ppq->p->screen, &res);
+   u_box_2d(0, 0, 165, 165, &box);
+
+   ppq->p->pipe->transfer_inline_write(ppq->p->pipe, areamaptex, 0,
+                                       PIPE_TRANSFER_WRITE, &box,
+                                       areamap, 165 * 2, sizeof(areamap));
+
+
+
+   ppq->shaders[n][1] = pp_tgsi_to_state(ppq->p->pipe, offsetvs, true,
+                                         "offsetvs");
+   if (iscolor)
+      ppq->shaders[n][2] = pp_tgsi_to_state(ppq->p->pipe, color1fs,
+                                            false, "color1fs");
+   else
+      ppq->shaders[n][2] = pp_tgsi_to_state(ppq->p->pipe, depth1fs,
+                                            false, "depth1fs");
+   ppq->shaders[n][3] = pp_tgsi_to_state(ppq->p->pipe, tmp_text, false,
+                                         "blend2fs");
+   ppq->shaders[n][4] = pp_tgsi_to_state(ppq->p->pipe, neigh3fs, false,
+                                         "neigh3fs");
+
+   free(tmp_text);
+}
+
+/** Short wrapper to init the depth version. */
+void
+pp_jimenezmlaa_init(struct pp_queue_t *ppq, unsigned int n, unsigned int val)
+{
+
+   pp_jimenezmlaa_init_run(ppq, n, val, false);
+}
+
+/** Short wrapper to init the color version. */
+void
+pp_jimenezmlaa_init_color(struct pp_queue_t *ppq, unsigned int n,
+                          unsigned int val)
+{
+
+   pp_jimenezmlaa_init_run(ppq, n, val, true);
+}
+
+/** Short wrapper to run the depth version. */
+void
+pp_jimenezmlaa(struct pp_queue_t *ppq, struct pipe_resource *in,
+               struct pipe_resource *out, unsigned int n)
+{
+   pp_jimenezmlaa_run(ppq, in, out, n, false);
+}
+
+/** Short wrapper to run the color version. */
+void
+pp_jimenezmlaa_color(struct pp_queue_t *ppq, struct pipe_resource *in,
+                     struct pipe_resource *out, unsigned int n)
+{
+   pp_jimenezmlaa_run(ppq, in, out, n, true);
+}
diff --git a/src/gallium/auxiliary/postprocess/pp_mlaa.h b/src/gallium/auxiliary/postprocess/pp_mlaa.h
new file mode 100644
index 00000000000..9972d59c6a6
--- /dev/null
+++ b/src/gallium/auxiliary/postprocess/pp_mlaa.h
@@ -0,0 +1,342 @@
+/**
+ * Copyright (C) 2010 Jorge Jimenez (jorge@iryoku.com)
+ * Copyright (C) 2010 Belen Masia (bmasia@unizar.es)
+ * Copyright (C) 2010 Jose I. Echevarria (joseignacioechevarria@gmail.com)
+ * Copyright (C) 2010 Fernando Navarro (fernandn@microsoft.com)
+ * Copyright (C) 2010 Diego Gutierrez (diegog@unizar.es)
+ * Copyright (C) 2011 Lauri Kasanen (cand@gmx.com)
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ *    1. Redistributions of source code must retain the above copyright notice,
+ *       this list of conditions and the following disclaimer.
+ *
+ *    2. Redistributions in binary form must reproduce the following statement:
+ *
+ *       "Uses Jimenez's MLAA. Copyright (C) 2010 by Jorge Jimenez, Belen Masia,
+ *        Jose I. Echevarria, Fernando Navarro and Diego Gutierrez."
+ *
+ *       Only for use in the Mesa project, this point 2 is filled by naming the
+ *       technique Jimenez's MLAA in the Mesa config options.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS
+ * IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
+ * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL COPYRIGHT HOLDERS OR CONTRIBUTORS
+ * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ *
+ * The views and conclusions contained in the software and documentation are
+ * those of the authors and should not be interpreted as representing official
+ * policies, either expressed or implied, of the copyright holders.
+ */
+
+#ifndef PP_MLAA_H
+#define PP_MLAA_H
+
+#include "postprocess/pp_mlaa_areamap.h"
+
+static const char depth1fs[] = "FRAG\n"
+   "PROPERTY FS_COLOR0_WRITES_ALL_CBUFS 1\n"
+   "DCL IN[0], GENERIC[0], PERSPECTIVE\n"
+   "DCL IN[1], GENERIC[10], PERSPECTIVE\n"
+   "DCL IN[2], GENERIC[11], PERSPECTIVE\n"
+   "DCL OUT[0], COLOR\n"
+   "DCL SAMP[0]\n"
+   "DCL TEMP[0..2]\n"
+   "IMM FLT32 {    0.0030,     0.0000,     1.0000,     0.0000}\n"
+   "  0: TEX TEMP[0].x, IN[1].xyyy, SAMP[0], 2D\n"
+   "  1: MOV TEMP[1].x, TEMP[0].xxxx\n"
+   "  2: TEX TEMP[0].x, IN[1].zwww, SAMP[0], 2D\n"
+   "  3: MOV TEMP[1].y, TEMP[0].xxxx\n"
+   "  4: TEX TEMP[0].x, IN[2].xyyy, SAMP[0], 2D\n"
+   "  5: MOV TEMP[1].z, TEMP[0].xxxx\n"
+   "  6: TEX TEMP[0].x, IN[2].zwww, SAMP[0], 2D\n"
+   "  7: MOV TEMP[1].w, TEMP[0].xxxx\n"
+   "  8: TEX TEMP[0].x, IN[0].xyyy, SAMP[0], 2D\n"
+   "  9: ADD TEMP[2], TEMP[0].xxxx, -TEMP[1]\n"
+   " 10: ABS TEMP[0], TEMP[2]\n"
+   " 11: SGE TEMP[2], TEMP[0], IMM[0].xxxx\n"
+   " 12: DP4 TEMP[0].x, TEMP[2], IMM[0].zzzz\n"
+   " 13: SEQ TEMP[1].x, TEMP[0].xxxx, IMM[0].yyyy\n"
+   " 14: IF TEMP[1].xxxx :16\n"
+   " 15:   KILP\n"
+   " 16: ENDIF\n"
+   " 17: MOV OUT[0], TEMP[2]\n"
+   " 18: END\n";
+
+
+static const char color1fs[] = "FRAG\n"
+   "PROPERTY FS_COLOR0_WRITES_ALL_CBUFS 1\n"
+   "DCL IN[0], GENERIC[0], PERSPECTIVE\n"
+   "DCL IN[1], GENERIC[10], PERSPECTIVE\n"
+   "DCL IN[2], GENERIC[11], PERSPECTIVE\n"
+   "DCL OUT[0], COLOR\n"
+   "DCL SAMP[0]\n"
+   "DCL TEMP[0..2]\n"
+   "IMM FLT32 {    0.2126,     0.7152,     0.0722,     0.1000}\n"
+   "IMM FLT32 {    1.0000,     0.0000,     0.0000,     0.0000}\n"
+   "  0: TEX TEMP[1].xyz, IN[1].xyyy, SAMP[0], 2D\n"
+   "  1: DP3 TEMP[0].x, TEMP[1].xyzz, IMM[0]\n"
+   "  2: TEX TEMP[1].xyz, IN[1].zwww, SAMP[0], 2D\n"
+   "  3: DP3 TEMP[0].y, TEMP[1].xyzz, IMM[0].xyzz\n"
+   "  4: TEX TEMP[1].xyz, IN[2].xyyy, SAMP[0], 2D\n"
+   "  5: DP3 TEMP[0].z, TEMP[1].xyzz, IMM[0].xyzz\n"
+   "  6: TEX TEMP[1].xyz, IN[2].zwww, SAMP[0], 2D\n"
+   "  7: DP3 TEMP[0].w, TEMP[1].xyzz, IMM[0].xyzz\n"
+   "  8: TEX TEMP[1].xyz, IN[0].xyyy, SAMP[0], 2D\n"
+   "  9: DP3 TEMP[2].x, TEMP[1].xyzz, IMM[0].xyzz\n"
+   " 10: ADD TEMP[1], TEMP[2].xxxx, -TEMP[0]\n"
+   " 11: ABS TEMP[0], TEMP[1]\n"
+   " 12: SGE TEMP[2], TEMP[0], IMM[0].wwww\n"
+   " 13: DP4 TEMP[0].x, TEMP[2], IMM[1].xxxx\n"
+   " 14: SEQ TEMP[1].x, TEMP[0].xxxx, IMM[1].yyyy\n"
+   " 15: IF TEMP[1].xxxx :17\n"
+   " 16:   KILP\n"
+   " 17: ENDIF\n"
+   " 18: MOV OUT[0], TEMP[2]\n"
+   " 19: END\n";
+
+
+static const char neigh3fs[] = "FRAG\n"
+   "PROPERTY FS_COLOR0_WRITES_ALL_CBUFS 1\n"
+   "DCL IN[0], GENERIC[0], PERSPECTIVE\n"
+   "DCL IN[1], GENERIC[10], PERSPECTIVE\n"
+   "DCL IN[2], GENERIC[11], PERSPECTIVE\n"
+   "DCL OUT[0], COLOR\n"
+   "DCL SAMP[0]\n"
+   "DCL SAMP[1]\n"
+   "DCL TEMP[0..8]\n"
+   "IMM FLT32 {    1.0000,     0.00001,     0.0000,     0.0000}\n"
+   "  0: TEX TEMP[0], IN[0].xyyy, SAMP[1], 2D\n"
+   "  1: MOV TEMP[1].x, TEMP[0].xxxx\n"
+   "  2: TEX TEMP[2].y, IN[2].zwww, SAMP[1], 2D\n"
+   "  3: MOV TEMP[1].y, TEMP[2].yyyy\n"
+   "  4: MOV TEMP[1].z, TEMP[0].zzzz\n"
+   "  5: TEX TEMP[1].w, IN[2].xyyy, SAMP[1], 2D\n"
+   "  6: MUL TEMP[4], TEMP[1], TEMP[1]\n"
+   "  7: MUL TEMP[5], TEMP[4], TEMP[1]\n"
+   "  8: DP4 TEMP[1].x, TEMP[5], IMM[0].xxxx\n"
+   "  9: SLT TEMP[4].x, TEMP[1].xxxx, IMM[0].yyyy\n"
+   " 10: IF TEMP[4].xxxx :12\n"
+   " 11:   KILP\n"
+   " 12: ENDIF\n"
+   " 13: TEX TEMP[4], IN[0].xyyy, SAMP[0], 2D\n"
+   " 14: TEX TEMP[6], IN[1].zwww, SAMP[0], 2D\n"
+   " 15: ADD TEMP[7].x, IMM[0].xxxx, -TEMP[0].xxxx\n"
+   " 16: MUL TEMP[8], TEMP[4], TEMP[7].xxxx\n"
+   " 17: MAD TEMP[7], TEMP[6], TEMP[0].xxxx, TEMP[8]\n"
+   " 18: MUL TEMP[6], TEMP[7], TEMP[5].xxxx\n"
+   " 19: TEX TEMP[7], IN[2].zwww, SAMP[0], 2D\n"
+   " 20: ADD TEMP[8].x, IMM[0].xxxx, -TEMP[2].yyyy\n"
+   " 21: MUL TEMP[3], TEMP[4], TEMP[8].xxxx\n"
+   " 22: MAD TEMP[8], TEMP[7], TEMP[2].yyyy, TEMP[3]\n"
+   " 23: MAD TEMP[2], TEMP[8], TEMP[5].yyyy, TEMP[6]\n"
+   " 24: TEX TEMP[6], IN[1].xyyy, SAMP[0], 2D\n"
+   " 25: ADD TEMP[7].x, IMM[0].xxxx, -TEMP[0].zzzz\n"
+   " 26: MUL TEMP[8], TEMP[4], TEMP[7].xxxx\n"
+   " 27: MAD TEMP[7], TEMP[6], TEMP[0].zzzz, TEMP[8]\n"
+   " 28: MAD TEMP[0], TEMP[7], TEMP[5].zzzz, TEMP[2]\n"
+   " 29: TEX TEMP[2], IN[2].xyyy, SAMP[0], 2D\n"
+   " 30: ADD TEMP[6].x, IMM[0].xxxx, -TEMP[1].wwww\n"
+   " 31: MUL TEMP[7], TEMP[4], TEMP[6].xxxx\n"
+   " 32: MAD TEMP[4], TEMP[2], TEMP[1].wwww, TEMP[7]\n"
+   " 33: MAD TEMP[2], TEMP[4], TEMP[5].wwww, TEMP[0]\n"
+   " 34: RCP TEMP[0].x, TEMP[1].xxxx\n"
+   " 35: MUL OUT[0], TEMP[2], TEMP[0].xxxx\n"
+   " 36: END\n";
+
+
+static const char offsetvs[] = "VERT\n"
+   "DCL IN[0]\n"
+   "DCL IN[1]\n"
+   "DCL OUT[0], POSITION\n"
+   "DCL OUT[1], GENERIC[0]\n"
+   "DCL OUT[2], GENERIC[10]\n"
+   "DCL OUT[3], GENERIC[11]\n"
+   "DCL CONST[0]\n"
+   "IMM FLT32 {    1.0000,     0.0000,    -1.0000,     0.0000}\n"
+   "  0: MOV OUT[0], IN[0]\n"
+   "  1: MOV OUT[1], IN[1]\n"
+   "  2: MAD OUT[2], CONST[0].xyxy, IMM[0].zyyz, IN[1].xyxy\n"
+   "  3: MAD OUT[3], CONST[0].xyxy, IMM[0].xyyx, IN[1].xyxy\n"
+   "  4: END\n";
+
+
+static const char blend2fs_1[] = "FRAG\n"
+   "PROPERTY FS_COLOR0_WRITES_ALL_CBUFS 1\n"
+   "DCL IN[0], GENERIC[0], PERSPECTIVE\n"
+   "DCL OUT[0], COLOR\n"
+   "DCL SAMP[0]\n"
+   "DCL SAMP[1]\n"
+   "DCL SAMP[2]\n"
+   "DCL CONST[0]\n"
+   "DCL TEMP[0..6]\n"
+   "IMM FLT32 {    0.0000,    -0.2500,     0.00609756,     0.5000}\n"
+   "IMM FLT32 {   -1.5000,    -2.0000,     0.9000,     1.5000}\n"
+   "IMM FLT32 {    2.0000,     1.0000,     4.0000,    33.0000}\n";
+
+static const char blend2fs_2[] =
+   "  0: MOV TEMP[0], IMM[0].xxxx\n"
+   "  1: TEX TEMP[1], IN[0].xyyy, SAMP[1], 2D\n"
+   "  2: MOV TEMP[2].x, TEMP[1]\n"
+   "  3: SNE TEMP[3].x, TEMP[1].yyyy, IMM[0].xxxx\n"
+   "  4: IF TEMP[3].xxxx :76\n"
+   "  5:   MOV TEMP[1].xy, IN[0].xyxx\n"
+   "  6:   MOV TEMP[4].x, IMM[1].xxxx\n"
+   "  7:   BGNLOOP :24\n"
+   "  8:     MUL TEMP[5].x, IMM[1].yyyy, IMM[3].xxxx\n"
+   "  9:     SLE TEMP[6].x, TEMP[4].xxxx, TEMP[5].xxxx\n"
+   " 10:     IF TEMP[6].xxxx :12\n"
+   " 11:       BRK\n"
+   " 12:     ENDIF\n"
+   " 13:     MOV TEMP[4].y, IMM[0].xxxx\n"
+   " 14:     MAD TEMP[3].xyz, CONST[0].xyyy, TEMP[4].xyyy, TEMP[1].xyyy\n"
+   " 15:     MOV TEMP[3].w, IMM[0].xxxx\n"
+   " 16:     TXL TEMP[5], TEMP[3], SAMP[2], 2D\n"
+   " 17:     MOV TEMP[3].x, TEMP[5].yyyy\n"
+   " 18:     SLT TEMP[6].x, TEMP[5].yyyy, IMM[1].zzzz\n"
+   " 19:     IF TEMP[6].xxxx :21\n"
+   " 20:       BRK\n"
+   " 21:     ENDIF\n"
+   " 22:     ADD TEMP[6].x, TEMP[4].xxxx, IMM[1].yyyy\n"
+   " 23:     MOV TEMP[4].x, TEMP[6].xxxx\n"
+   " 24:   ENDLOOP :7\n"
+   " 25:   ADD TEMP[1].x, TEMP[4].xxxx, IMM[1].wwww\n"
+   " 26:   MAD TEMP[6].x, -IMM[2].xxxx, TEMP[3].xxxx, TEMP[1].xxxx\n"
+   " 27:   MUL TEMP[1].x, IMM[1].yyyy, IMM[3].xxxx\n"
+   " 28:   MAX TEMP[4].x, TEMP[6].xxxx, TEMP[1].xxxx\n"
+   " 29:   MOV TEMP[1].x, TEMP[4].xxxx\n"
+   " 30:   MOV TEMP[3].xy, IN[0].xyxx\n"
+   " 31:   MOV TEMP[5].x, IMM[1].wwww\n"
+   " 32:   BGNLOOP :49\n"
+   " 33:     MUL TEMP[6].x, IMM[2].xxxx, IMM[3].xxxx\n"
+   " 34:     SGE TEMP[4].x, TEMP[5].xxxx, TEMP[6].xxxx\n"
+   " 35:     IF TEMP[4].xxxx :37\n"
+   " 36:       BRK\n"
+   " 37:     ENDIF\n"
+   " 38:     MOV TEMP[5].y, IMM[0].xxxx\n"
+   " 39:     MAD TEMP[4].xyz, CONST[0].xyyy, TEMP[5].xyyy, TEMP[3].xyyy\n"
+   " 40:     MOV TEMP[4].w, IMM[0].xxxx\n"
+   " 41:     TXL TEMP[6].xy, TEMP[4], SAMP[2], 2D\n"
+   " 42:     MOV TEMP[4].x, TEMP[6].yyyy\n"
+   " 43:     SLT TEMP[0].x, TEMP[6].yyyy, IMM[1].zzzz\n"
+   " 44:     IF TEMP[0].xxxx :46\n"
+   " 45:       BRK\n"
+   " 46:     ENDIF\n"
+   " 47:     ADD TEMP[6].x, TEMP[5].xxxx, IMM[2].xxxx\n"
+   " 48:     MOV TEMP[5].x, TEMP[6].xxxx\n"
+   " 49:   ENDLOOP :32\n"
+   " 50:   ADD TEMP[3].x, TEMP[5].xxxx, IMM[1].xxxx\n"
+   " 51:   MAD TEMP[5].x, IMM[2].xxxx, TEMP[4].xxxx, TEMP[3].xxxx\n"
+   " 52:   MUL TEMP[3].x, IMM[2].xxxx, IMM[3].xxxx\n"
+   " 53:   MIN TEMP[4].x, TEMP[5].xxxx, TEMP[3].xxxx\n"
+   " 54:   MOV TEMP[3].x, TEMP[1].xxxx\n"
+   " 55:   MOV TEMP[3].y, TEMP[4].xxxx\n"
+   " 56:   MOV TEMP[5].yw, IMM[0].yyyy\n"
+   " 57:   MOV TEMP[5].x, TEMP[1].xxxx\n"
+   " 58:   ADD TEMP[1].x, TEMP[4].xxxx, IMM[2].yyyy\n"
+   " 59:   MOV TEMP[5].z, TEMP[1].xxxx\n"
+   " 60:   MAD TEMP[1], TEMP[5], CONST[0].xyxy, IN[0].xyxy\n"
+   " 61:   MOV TEMP[4], TEMP[1].xyyy\n"
+   " 62:   MOV TEMP[4].w, IMM[0].xxxx\n"
+   " 63:   TXL TEMP[5].x, TEMP[4], SAMP[2], 2D\n"
+   " 64:   MOV TEMP[4].x, TEMP[5].xxxx\n"
+   " 65:   MOV TEMP[5], TEMP[1].zwww\n"
+   " 66:   MOV TEMP[5].w, IMM[0].xxxx\n"
+   " 67:   TXL TEMP[1].x, TEMP[5], SAMP[2], 2D\n"
+   " 68:   MOV TEMP[4].y, TEMP[1].xxxx\n"
+   " 69:   MUL TEMP[5].xy, IMM[2].zzzz, TEMP[4].xyyy\n"
+   " 70:   ROUND TEMP[1].xy, TEMP[5].xyyy\n"
+   " 71:   ABS TEMP[4].xy, TEMP[3].xyyy\n"
+   " 72:   MAD TEMP[3].xy, IMM[2].wwww, TEMP[1].xyyy, TEMP[4].xyyy\n"
+   " 73:   MUL TEMP[5].xyz, TEMP[3].xyyy, IMM[0].zzzz\n"
+   " 74:   MOV TEMP[5].w, IMM[0].xxxx\n"
+   " 75:   TXL TEMP[0].xy, TEMP[5], SAMP[0], 2D\n"
+   " 76: ENDIF\n"
+   " 77: SNE TEMP[1].x, TEMP[2].xxxx, IMM[0].xxxx\n"
+   " 78: IF TEMP[1].xxxx :151\n"
+   " 79:   MOV TEMP[1].xy, IN[0].xyxx\n"
+   " 80:   MOV TEMP[3].x, IMM[1].xxxx\n"
+   " 81:   BGNLOOP :98\n"
+   " 82:     MUL TEMP[4].x, IMM[1].yyyy, IMM[3].xxxx\n"
+   " 83:     SLE TEMP[5].x, TEMP[3].xxxx, TEMP[4].xxxx\n"
+   " 84:     IF TEMP[5].xxxx :86\n"
+   " 85:       BRK\n"
+   " 86:     ENDIF\n"
+   " 87:     MOV TEMP[3].y, IMM[0].xxxx\n"
+   " 88:     MAD TEMP[5].xyz, CONST[0].xyyy, TEMP[3].yxxx, TEMP[1].xyyy\n"
+   " 89:     MOV TEMP[5].w, IMM[0].xxxx\n"
+   " 90:     TXL TEMP[4], TEMP[5], SAMP[2], 2D\n"
+   " 91:     MOV TEMP[2].x, TEMP[4].xxxx\n"
+   " 92:     SLT TEMP[5].x, TEMP[4].xxxx, IMM[1].zzzz\n"
+   " 93:     IF TEMP[5].xxxx :95\n"
+   " 94:       BRK\n"
+   " 95:     ENDIF\n"
+   " 96:     ADD TEMP[4].x, TEMP[3].xxxx, IMM[1].yyyy\n"
+   " 97:     MOV TEMP[3].x, TEMP[4].xxxx\n"
+   " 98:   ENDLOOP :81\n"
+   " 99:   ADD TEMP[1].x, TEMP[3].xxxx, IMM[1].wwww\n"
+   "100:   MAD TEMP[6].x, -IMM[2].xxxx, TEMP[2].xxxx, TEMP[1].xxxx\n"
+   "101:   MUL TEMP[1].x, IMM[1].yyyy, IMM[3].xxxx\n"
+   "102:   MAX TEMP[3].x, TEMP[6].xxxx, TEMP[1].xxxx\n"
+   "103:   MOV TEMP[1].x, TEMP[3].xxxx\n"
+   "104:   MOV TEMP[2].xy, IN[0].xyxx\n"
+   "105:   MOV TEMP[4].x, IMM[1].wwww\n"
+   "106:   BGNLOOP :123\n"
+   "107:     MUL TEMP[5].x, IMM[2].xxxx, IMM[3].xxxx\n"
+   "108:     SGE TEMP[6].x, TEMP[4].xxxx, TEMP[5].xxxx\n"
+   "109:     IF TEMP[6].xxxx :111\n"
+   "110:       BRK\n"
+   "111:     ENDIF\n"
+   "112:     MOV TEMP[4].y, IMM[0].xxxx\n"
+   "113:     MAD TEMP[5].xyz, CONST[0].xyyy, TEMP[4].yxxx, TEMP[2].xyyy\n"
+   "114:     MOV TEMP[5].w, IMM[0].xxxx\n"
+   "115:     TXL TEMP[6], TEMP[5], SAMP[2], 2D\n"
+   "116:     MOV TEMP[3].x, TEMP[6].xxxx\n"
+   "117:     SLT TEMP[5].x, TEMP[6].xxxx, IMM[1].zzzz\n"
+   "118:     IF TEMP[5].xxxx :120\n"
+   "119:       BRK\n"
+   "120:     ENDIF\n"
+   "121:     ADD TEMP[6].x, TEMP[4].xxxx, IMM[2].xxxx\n"
+   "122:     MOV TEMP[4].x, TEMP[6].xxxx\n"
+   "123:   ENDLOOP :106\n"
+   "124:   ADD TEMP[2].x, TEMP[4].xxxx, IMM[1].xxxx\n"
+   "125:   MAD TEMP[4].x, IMM[2].xxxx, TEMP[3].xxxx, TEMP[2].xxxx\n"
+   "126:   MUL TEMP[2].x, IMM[2].xxxx, IMM[3].xxxx\n"
+   "127:   MIN TEMP[3].x, TEMP[4].xxxx, TEMP[2].xxxx\n"
+   "128:   MOV TEMP[2].x, TEMP[1].xxxx\n"
+   "129:   MOV TEMP[2].y, TEMP[3].xxxx\n"
+   "130:   MOV TEMP[4].xz, IMM[0].yyyy\n"
+   "131:   MOV TEMP[4].y, TEMP[1].xxxx\n"
+   "132:   ADD TEMP[1].x, TEMP[3].xxxx, IMM[2].yyyy\n"
+   "133:   MOV TEMP[4].w, TEMP[1].xxxx\n"
+   "134:   MAD TEMP[1], TEMP[4], CONST[0].xyxy, IN[0].xyxy\n"
+   "135:   MOV TEMP[3], TEMP[1].xyyy\n"
+   "136:   MOV TEMP[3].w, IMM[0].xxxx\n"
+   "137:   TXL TEMP[4].y, TEMP[3], SAMP[2], 2D\n"
+   "138:   MOV TEMP[3].x, TEMP[4].yyyy\n"
+   "139:   MOV TEMP[4], TEMP[1].zwww\n"
+   "140:   MOV TEMP[4].w, IMM[0].xxxx\n"
+   "141:   TXL TEMP[1].y, TEMP[4], SAMP[2], 2D\n"
+   "142:   MOV TEMP[3].y, TEMP[1].yyyy\n"
+   "143:   MUL TEMP[4].xy, IMM[2].zzzz, TEMP[3].xyyy\n"
+   "144:   ROUND TEMP[1].xy, TEMP[4].xyyy\n"
+   "145:   ABS TEMP[3].xy, TEMP[2].xyyy\n"
+   "146:   MAD TEMP[2].xy, IMM[2].wwww, TEMP[1].xyyy, TEMP[3].xyyy\n"
+   "147:   MUL TEMP[3].xyz, TEMP[2].xyyy, IMM[0].zzzz\n"
+   "148:   MOV TEMP[3].w, IMM[0].xxxx\n"
+   "149:   TXL TEMP[1].xy, TEMP[3], SAMP[0], 2D\n"
+   "150:   MOV TEMP[0].zw, TEMP[1].yyxy\n"
+   "151: ENDIF\n"
+   "152: MOV OUT[0], TEMP[0]\n"
+   "153: END\n";
+
+#endif