From: Brian Paul <brianp@vmware.com>
Date: Fri, 12 Jan 2018 20:18:25 +0000 (-0700)
Subject: vbo: optimize some display list drawing (v2)
X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=8e4efdc895eacca931b94a0c1fa01aae5c34463b;p=mesa.git

vbo: optimize some display list drawing (v2)

The vbo_save_vertex_list structure records one or more glBegin/End
primitives which all have the same vertex format.

To draw these primitives, we setup the vertex array state, then
issue the drawing command.  Before, the 'start' vertex was typically
zero and we used the vertex array pointer to indicate where the
vertex data starts.

This patch checks if the vertex buffer offset is an exact multiple of
the vertex size.  If so, that means we can use zero-based vertex array
pointers and use the draw's start value to indicate where the vertex
data starts.

This means a series of display list drawing commands may have
identical vertex array state.  This will get filtered out by the
Gallium CSO module so we can issue a tight series of drawing commands
without state changes to the device.

Note that this also works for a series of glCallList commands (not
just one list that contains multiple glBegin/End pairs).

No Piglit or conform changes.

v2: minor fixes suggested by Ian.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
---

diff --git a/src/mesa/vbo/vbo_save.h b/src/mesa/vbo/vbo_save.h
index 9d13e0aa18d..274d667be78 100644
--- a/src/mesa/vbo/vbo_save.h
+++ b/src/mesa/vbo/vbo_save.h
@@ -86,6 +86,20 @@ struct vbo_save_vertex_list {
    struct vbo_save_primitive_store *prim_store;
 };
 
+
+/**
+ * Is the vertex lists's buffer offset an exact multiple of the
+ * vertex size (in bytes)?  This is used to check for a vertex array /
+ * drawing optimization.
+ */
+static inline bool
+aligned_vertex_buffer_offset(const struct vbo_save_vertex_list *node)
+{
+   unsigned vertex_size = node->vertex_size * sizeof(GLfloat); /* in bytes */
+   return vertex_size != 0 && node->buffer_offset % vertex_size == 0;
+}
+
+
 /* These buffers should be a reasonable size to support upload to
  * hardware.  Current vbo implementation will re-upload on any
  * changes, so don't make too big or apps which dynamically create
diff --git a/src/mesa/vbo/vbo_save_api.c b/src/mesa/vbo/vbo_save_api.c
index 42d883fed4d..1c575448773 100644
--- a/src/mesa/vbo/vbo_save_api.c
+++ b/src/mesa/vbo/vbo_save_api.c
@@ -546,6 +546,20 @@ compile_vertex_list(struct gl_context *ctx)
       save->prim_store = alloc_prim_store();
    }
 
+   /*
+    * If the vertex buffer offset is a multiple of the vertex size,
+    * we can use the _mesa_prim::start value to indicate where the
+    * vertices starts, instead of the buffer offset.  Also see the
+    * bind_vertex_list() function.
+    */
+   if (aligned_vertex_buffer_offset(node)) {
+      const unsigned start_offset =
+         node->buffer_offset / (node->vertex_size * sizeof(GLfloat));
+      for (unsigned i = 0; i < save->prim_count; i++) {
+         save->prims[i].start += start_offset;
+      }
+   }
+
    /* Reset our structures for the next run of vertices:
     */
    reset_counters(ctx);
diff --git a/src/mesa/vbo/vbo_save_draw.c b/src/mesa/vbo/vbo_save_draw.c
index 1694a04a3e1..b63a9a8b869 100644
--- a/src/mesa/vbo/vbo_save_draw.c
+++ b/src/mesa/vbo/vbo_save_draw.c
@@ -146,6 +146,18 @@ bind_vertex_list(struct gl_context *ctx,
    memcpy(node_attrsz, node->attrsz, sizeof(node->attrsz));
    memcpy(node_attrtype, node->attrtype, sizeof(node->attrtype));
 
+   if (aligned_vertex_buffer_offset(node)) {
+      /* The vertex size is an exact multiple of the buffer offset.
+       * This means that we can use zero-based vertex attribute pointers
+       * and specify the start of the primitive with the _mesa_prim::start
+       * field.  This results in issuing several draw calls with identical
+       * vertex attribute information.  This can result in fewer state
+       * changes in drivers.  In particular, the Gallium CSO module will
+       * filter out redundant vertex buffer changes.
+       */
+      buffer_offset = 0;
+   }
+
    /* Install the default (ie Current) attributes first, then overlay
     * all active ones.
     */