translate_sse: remove useless generated function wrappers

[mesa.git] / src / gallium / auxiliary / translate / translate_sse.c

diff --git a/src/gallium/auxiliary/translate/translate_sse.c b/src/gallium/auxiliary/translate/translate_sse.c
index ba4a246fdbf1c4f8e48b7983526b981ddc64d918..68c71f42513357e5a7d3ac988114146f12604a4e 100644 (file)
--- a/src/gallium/auxiliary/translate/translate_sse.c
+++ b/src/gallium/auxiliary/translate/translate_sse.c
@@ -46,21 +46,10 @@
 #define W    3
 
 
-typedef void (PIPE_CDECL *run_func)( struct translate *translate,
-                                     unsigned start,
-                                     unsigned count,
-                                     unsigned instance_id,
-                                     void *output_buffer );
-
-typedef void (PIPE_CDECL *run_elts_func)( struct translate *translate,
-                                          const unsigned *elts,
-                                          unsigned count,
-                                          unsigned instance_id,
-                                          void *output_buffer );
-
 struct translate_buffer {
    const void *base_ptr;
    unsigned stride;
+   unsigned max_index;
 };
 
 struct translate_buffer_varient {
@@ -70,6 +59,9 @@ struct translate_buffer_varient {
 };
 
 
+#define ELEMENT_BUFFER_INSTANCE_ID  1001
+
+
 struct translate_sse {
    struct translate translate;
 
@@ -98,9 +90,6 @@ struct translate_sse {
    boolean use_instancing;
    unsigned instance_id;
 
-   run_func      gen_run;
-   run_elts_func gen_run_elts;
-
    /* these are actually known values, but putting them in a struct
     * like this is helpful to keep them in sync across the file.
     */
@@ -420,6 +409,11 @@ static boolean init_inputs( struct translate_sse *p,
          } else {
             x86_mov(p->func, tmp_EAX, elt);
          }
+
+         /*
+          * TODO: Respect translate_buffer::max_index.
+          */
+
          x86_imul(p->func, tmp_EAX, buf_stride);
          x86_add(p->func, tmp_EAX, buf_base_ptr);
 
@@ -443,6 +437,10 @@ static struct x86_reg get_buffer_ptr( struct translate_sse *p,
                                       unsigned var_idx,
                                       struct x86_reg elt )
 {
+   if (var_idx == ELEMENT_BUFFER_INSTANCE_ID) {
+      return x86_make_disp(p->machine_EDX,
+                           get_offset(p, &p->instance_id));
+   }
    if (linear && p->nr_buffer_varients == 1) {
       return p->idx_EBX;
    }
@@ -659,13 +657,15 @@ static boolean build_vertex_emit( struct translate_sse *p,
 static void translate_sse_set_buffer( struct translate *translate,
                                unsigned buf,
                                const void *ptr,
-                               unsigned stride )
+                               unsigned stride,
+                               unsigned max_index )
 {
    struct translate_sse *p = (struct translate_sse *)translate;
 
    if (buf < p->nr_buffers) {
       p->buffer[buf].base_ptr = (char *)ptr;
       p->buffer[buf].stride = stride;
+      p->buffer[buf].max_index = max_index;
    }
 
    if (0) debug_printf("%s %d/%d: %p %d\n", 
@@ -685,36 +685,6 @@ static void translate_sse_release( struct translate *translate )
    FREE(p);
 }
 
-static void PIPE_CDECL translate_sse_run_elts( struct translate *translate,
-                             const unsigned *elts,
-                             unsigned count,
-                              unsigned instance_id,
-                             void *output_buffer )
-{
-   struct translate_sse *p = (struct translate_sse *)translate;
-
-   p->gen_run_elts( translate,
-                   elts,
-                   count,
-                    instance_id,
-                   output_buffer );
-}
-
-static void PIPE_CDECL translate_sse_run( struct translate *translate,
-                        unsigned start,
-                        unsigned count,
-                         unsigned instance_id,
-                        void *output_buffer )
-{
-   struct translate_sse *p = (struct translate_sse *)translate;
-
-   p->gen_run( translate,
-              start,
-              count,
-               instance_id,
-              output_buffer );
-}
-
 
 struct translate *translate_sse2_create( const struct translate_key *key )
 {
@@ -731,33 +701,37 @@ struct translate *translate_sse2_create( const struct translate_key *key )
    p->translate.key = *key;
    p->translate.release = translate_sse_release;
    p->translate.set_buffer = translate_sse_set_buffer;
-   p->translate.run_elts = translate_sse_run_elts;
-   p->translate.run = translate_sse_run;
 
    for (i = 0; i < key->nr_elements; i++) {
-      unsigned j;
+      if (key->element[i].type == TRANSLATE_ELEMENT_NORMAL) {
+         unsigned j;
 
-      p->nr_buffers = MAX2( p->nr_buffers, key->element[i].input_buffer + 1 );
+         p->nr_buffers = MAX2(p->nr_buffers, key->element[i].input_buffer + 1);
 
-      if (key->element[i].instance_divisor) {
-         p->use_instancing = TRUE;
-      }
+         if (key->element[i].instance_divisor) {
+            p->use_instancing = TRUE;
+         }
 
-      /*
-       * Map vertex element to vertex buffer varient.
-       */
-      for (j = 0; j < p->nr_buffer_varients; j++) {
-         if (p->buffer_varient[j].buffer_index == key->element[i].input_buffer &&
-             p->buffer_varient[j].instance_divisor == key->element[i].instance_divisor) {
-            break;
+         /*
+          * Map vertex element to vertex buffer varient.
+          */
+         for (j = 0; j < p->nr_buffer_varients; j++) {
+            if (p->buffer_varient[j].buffer_index == key->element[i].input_buffer &&
+                p->buffer_varient[j].instance_divisor == key->element[i].instance_divisor) {
+               break;
+            }
          }
+         if (j == p->nr_buffer_varients) {
+            p->buffer_varient[j].buffer_index = key->element[i].input_buffer;
+            p->buffer_varient[j].instance_divisor = key->element[i].instance_divisor;
+            p->nr_buffer_varients++;
+         }
+         p->element_to_buffer_varient[i] = j;
+      } else {
+         assert(key->element[i].type == TRANSLATE_ELEMENT_INSTANCE_ID);
+
+         p->element_to_buffer_varient[i] = ELEMENT_BUFFER_INSTANCE_ID;
       }
-      if (j == p->nr_buffer_varients) {
-         p->buffer_varient[j].buffer_index = key->element[i].input_buffer;
-         p->buffer_varient[j].instance_divisor = key->element[i].instance_divisor;
-         p->nr_buffer_varients++;
-      }
-      p->element_to_buffer_varient[i] = j;
    }
 
    if (0) debug_printf("nr_buffers: %d\n", p->nr_buffers);
@@ -768,12 +742,12 @@ struct translate *translate_sse2_create( const struct translate_key *key )
    if (!build_vertex_emit(p, &p->elt_func, FALSE))
       goto fail;
 
-   p->gen_run = (run_func)x86_get_func(&p->linear_func);
-   if (p->gen_run == NULL)
+   p->translate.run = (void*)x86_get_func(&p->linear_func);
+   if (p->translate.run == NULL)
       goto fail;
 
-   p->gen_run_elts = (run_elts_func)x86_get_func(&p->elt_func);
-   if (p->gen_run_elts == NULL)
+   p->translate.run_elts = (void*)x86_get_func(&p->elt_func);
+   if (p->translate.run_elts == NULL)
       goto fail;
 
    return &p->translate;