gallium/i915: overhaul of fragment shader compilation, constant/immediate allocation

Before, fragment shaders were translated to i915 hw code at bind time, rather
than create time.  Now there's an i915_fragment_shader struct with the expected
contents that's created by i915_create_fs_state().  Translation to i915 code
takes place there too.

Immediates are handled correctly now.  During program translation we keep
track of which constant buffer slots are free (i.e. not referenced by the
shader).  Then the TGSI immediates and ancillary immediates (introduced for
SIN/COS/etc) are put into those free slots.

When it's time to upload the constant buffer, use the fp->constant_flags[]
array to determine if we should grab an immediate from the shader, or a
user-defined parameter from the gallium constant buffer.

1 files changed, 31 insertions, 8 deletions

diff --git a/src/gallium/drivers/i915simple/i915_state_emit.c b/src/gallium/drivers/i915simple/i915_state_emit.c
index 3339287f49..6bbaac4e34 100644
--- a/src/gallium/drivers/i915simple/i915_state_emit.c
+++ b/src/gallium/drivers/i915simple/i915_state_emit.c
@@ -99,7 +99,11 @@ i915_emit_hardware_state(struct i915_context *i915 )
                              2 + I915_TEX_UNITS*3 + 
                              2 + I915_TEX_UNITS*3 +
                              2 + I915_MAX_CONSTANT*4 + 
+#if 0
                              i915->current.program_len + 
+#else
+                             i915->fs->program_len + 
+#endif
                              6 
                            ) * 3/2; /* plus 50% margin */
    const unsigned relocs = ( I915_TEX_UNITS +
@@ -325,15 +329,34 @@ i915_emit_hardware_state(struct i915_context *i915 )
    /* 2 + I915_MAX_CONSTANT*4 dwords, 0 relocs */
    if (i915->hardware_dirty & I915_HW_PROGRAM)
    {
-      const uint nr = i915->current.num_constants[PIPE_SHADER_FRAGMENT];
-      assert(nr <= I915_MAX_CONSTANT);
-      if (nr > 0) {
-         const uint *c
-            = (const uint *) i915->current.constants[PIPE_SHADER_FRAGMENT];
+      /* Collate the user-defined constants with the fragment shader's
+       * immediates according to the constant_flags[] array.
+       */
+      const uint nr = i915->fs->num_constants;
+      if (nr) {
          uint i;
+
          OUT_BATCH( _3DSTATE_PIXEL_SHADER_CONSTANTS | (nr * 4) );
          OUT_BATCH( (1 << (nr - 1)) | ((1 << (nr - 1)) - 1) );
+
          for (i = 0; i < nr; i++) {
+            const uint *c;
+            if (i915->fs->constant_flags[i] == I915_CONSTFLAG_USER) {
+               /* grab user-defined constant */
+               c = (uint *) i915->current.constants[PIPE_SHADER_FRAGMENT][i];
+            }
+            else {
+               /* emit program constant */
+               c = (uint *) i915->fs->constants[i];
+            }
+#if 0 /* debug */
+            {
+               float *f = (float *) c;
+               printf("Const %2d: %f %f %f %f %s\n", i, f[0], f[1], f[2], f[3],
+                      (i915->fs->constant_flags[i] == I915_CONSTFLAG_USER
+                       ? "user" : "immediate"));
+            }
+#endif
             OUT_BATCH(*c++);
             OUT_BATCH(*c++);
             OUT_BATCH(*c++);
@@ -348,9 +371,9 @@ i915_emit_hardware_state(struct i915_context *i915 )
    {
       uint i;
       /* we should always have, at least, a pass-through program */
-      assert(i915->current.program_len > 0);
-      for (i = 0; i < i915->current.program_len; i++) {
-         OUT_BATCH(i915->current.program[i]);
+      assert(i915->fs->program_len > 0);
+      for (i = 0; i < i915->fs->program_len; i++) {
+         OUT_BATCH(i915->fs->program[i]);
       }
    }