3D View: new nethod of opengl selection

Intended to replace legacy GL_SELECT, without the limitations of
sample queries which can't access depth information.

This commit adds VIEW3D_SELECT_PICK_NEAREST and VIEW3D_SELECT_PICK_ALL
which access the depth buffers to detect whats under the pointer,
so initial selection is always the closest item.

The performance of this method depends a lot on the OpenGL
implementations glReadPixels.

Since reading depth can be slow, buffers are cached for object picking
so selecting re-uses depth data, performing 1 draw instead of 3
(for 24, 18, 10 px regions, picking with many items under the pointer).

Occlusion queries draw twice when picking nearest,
so worst case 6x draw calls per selection.

Even with these improvements occlusion queries is faster on AMD hardware.

Depth selection is disabled by default, toggle option under select method.
May enable by default if this works well on different hardware.

Reviewed as D2543

1 files changed, 118 insertions, 149 deletions

diff --git a/source/blender/gpu/intern/gpu_select.c b/source/blender/gpu/intern/gpu_select.c
index 35944c455a5..9496ff137dc 100644
--- a/source/blender/gpu/intern/gpu_select.c
+++ b/source/blender/gpu/intern/gpu_select.c
@@ -29,106 +29,86 @@
  * Interface for accessing gpu-related methods for selection. The semantics will be
  * similar to glRenderMode(GL_SELECT) since the goal is to maintain compatibility.
  */
+#include <stdlib.h>
+
 #include "GPU_select.h"
 #include "GPU_extensions.h"
 #include "GPU_glew.h"
- 
+
 #include "MEM_guardedalloc.h"
 
 #include "DNA_userdef_types.h"
 
-#include "BLI_rect.h"
-
 #include "BLI_utildefines.h"
 
-/* Ad hoc number of queries to allocate to skip doing many glGenQueries */
-#define ALLOC_QUERIES 200
-
-typedef struct GPUQueryState {
+#include "gpu_select_private.h"
+
+/* Internal algorithm used */
+enum {
+	/** GL_SELECT, legacy OpenGL selection */
+	ALGO_GL_LEGACY = 1,
+	/** glBegin/EndQuery(GL_SAMPLES_PASSED... ), `gpu_select_query.c`
+	 * Only sets 4th component (ID) correctly. */
+	ALGO_GL_QUERY = 2,
+	/** Read depth buffer for every drawing pass and extract depths, `gpu_select_pick.c`
+	 * Only sets 4th component (ID) correctly. */
+	ALGO_GL_PICK = 3,
+};
+
+typedef struct GPUSelectState {
 	/* To ignore selection id calls when not initialized */
 	bool select_is_active;
-	/* Tracks whether a query has been issued so that gpu_load_id can end the previous one */
-	bool query_issued;
-	/* array holding the OpenGL query identifiers */
-	unsigned int *queries;
-	/* array holding the id corresponding to each query */
-	unsigned int *id;
-	/* number of queries in *queries and *id */
-	unsigned int num_of_queries;
-	/* index to the next query to start */
-	unsigned int active_query;
 	/* flag to cache user preference for occlusion based selection */
 	bool use_gpu_select;
-	/* cache on initialization */
-	unsigned int *buffer;
-	/* buffer size (stores number of integers, for actual size multiply by sizeof integer)*/
-	unsigned int bufsize;
 	/* mode of operation */
 	char mode;
-	unsigned int index;
-	int oldhits;
-} GPUQueryState;
+	/* internal algorithm for selection */
+	char algorithm;
+	/* allow GPU_select_begin/end without drawing */
+	bool use_cache;
+} GPUSelectState;
 
-static GPUQueryState g_query_state = {0};
+static GPUSelectState g_select_state = {0};
 
 /**
  * initialize and provide buffer for results
  */
 void GPU_select_begin(unsigned int *buffer, unsigned int bufsize, const rcti *input, char mode, int oldhits)
 {
-	g_query_state.select_is_active = true;
-	g_query_state.query_issued = false;
-	g_query_state.active_query = 0;
-	g_query_state.use_gpu_select = GPU_select_query_check_active();
-	g_query_state.num_of_queries = 0;
-	g_query_state.bufsize = bufsize;
-	g_query_state.buffer = buffer;
-	g_query_state.mode = mode;
-	g_query_state.index = 0;
-	g_query_state.oldhits = oldhits;
+	g_select_state.select_is_active = true;
+	g_select_state.use_gpu_select = GPU_select_query_check_active();
+	g_select_state.mode = mode;
 
-	if (!g_query_state.use_gpu_select) {
-		glSelectBuffer(bufsize, (GLuint *)buffer);
-		glRenderMode(GL_SELECT);
-		glInitNames();
-		glPushName(-1);
+	if (ELEM(g_select_state.mode, GPU_SELECT_PICK_ALL, GPU_SELECT_PICK_NEAREST)) {
+		g_select_state.algorithm = ALGO_GL_PICK;
+	}
+	else if (!g_select_state.use_gpu_select) {
+		g_select_state.algorithm = ALGO_GL_LEGACY;
 	}
 	else {
-		float viewport[4];
-
-		g_query_state.num_of_queries = ALLOC_QUERIES;
-
-		g_query_state.queries = MEM_mallocN(g_query_state.num_of_queries * sizeof(*g_query_state.queries), "gpu selection queries");
-		g_query_state.id = MEM_mallocN(g_query_state.num_of_queries * sizeof(*g_query_state.id), "gpu selection ids");
-		glGenQueries(g_query_state.num_of_queries, g_query_state.queries);
-
-		glPushAttrib(GL_DEPTH_BUFFER_BIT | GL_VIEWPORT_BIT);
-		/* disable writing to the framebuffer */
-		glColorMask(GL_FALSE, GL_FALSE, GL_FALSE, GL_FALSE);
-
-		/* In order to save some fill rate we minimize the viewport using rect.
-		 * We need to get the region of the scissor so that our geometry doesn't
-		 * get rejected before the depth test. Should probably cull rect against
-		 * scissor for viewport but this is a rare case I think */
-		glGetFloatv(GL_SCISSOR_BOX, viewport);
-		glViewport(viewport[0], viewport[1], BLI_rcti_size_x(input), BLI_rcti_size_y(input));
+		g_select_state.algorithm = ALGO_GL_QUERY;
+	}
 
-		/* occlusion queries operates on fragments that pass tests and since we are interested on all
-		 * objects in the view frustum independently of their order, we need to disable the depth test */
-		if (mode == GPU_SELECT_ALL) {
-			glDisable(GL_DEPTH_TEST);
-			glDepthMask(GL_FALSE);
+	switch (g_select_state.algorithm) {
+		case ALGO_GL_LEGACY:
+		{
+			g_select_state.use_cache = false;
+			glSelectBuffer(bufsize, (GLuint *)buffer);
+			glRenderMode(GL_SELECT);
+			glInitNames();
+			glPushName(-1);
+			break;
 		}
-		else if (mode == GPU_SELECT_NEAREST_FIRST_PASS) {
-			glClear(GL_DEPTH_BUFFER_BIT);
-			glEnable(GL_DEPTH_TEST);
-			glDepthMask(GL_TRUE);
-			glDepthFunc(GL_LEQUAL);
+		case ALGO_GL_QUERY:
+		{
+			g_select_state.use_cache = false;
+			gpu_select_query_begin((unsigned int (*)[4])buffer, bufsize / 4, input, mode, oldhits);
+			break;
 		}
-		else if (mode == GPU_SELECT_NEAREST_SECOND_PASS) {
-			glEnable(GL_DEPTH_TEST);
-			glDepthMask(GL_FALSE);
-			glDepthFunc(GL_EQUAL);
+		default:  /* ALGO_GL_PICK */
+		{
+			gpu_select_pick_begin((unsigned int (*)[4])buffer, bufsize / 4, input, mode);
+			break;
 		}
 	}
 }
@@ -143,41 +123,24 @@ void GPU_select_begin(unsigned int *buffer, unsigned int bufsize, const rcti *in
 bool GPU_select_load_id(unsigned int id)
 {
 	/* if no selection mode active, ignore */
-	if (!g_query_state.select_is_active)
+	if (!g_select_state.select_is_active)
 		return true;
 
-	if (!g_query_state.use_gpu_select) {
-		glLoadName(id);
-	}
-	else {
-		if (g_query_state.query_issued) {
-			glEndQuery(GL_SAMPLES_PASSED);
+	switch (g_select_state.algorithm) {
+		case ALGO_GL_LEGACY:
+		{
+			glLoadName(id);
+			return true;
 		}
-		/* if required, allocate extra queries */
-		if (g_query_state.active_query == g_query_state.num_of_queries) {
-			g_query_state.num_of_queries += ALLOC_QUERIES;
-			g_query_state.queries = MEM_reallocN(g_query_state.queries, g_query_state.num_of_queries * sizeof(*g_query_state.queries));
-			g_query_state.id = MEM_reallocN(g_query_state.id, g_query_state.num_of_queries * sizeof(*g_query_state.id));
-			glGenQueries(ALLOC_QUERIES, &g_query_state.queries[g_query_state.active_query]);
+		case ALGO_GL_QUERY:
+		{
+			return gpu_select_query_load_id(id);
 		}
-
-		glBeginQuery(GL_SAMPLES_PASSED, g_query_state.queries[g_query_state.active_query]);
-		g_query_state.id[g_query_state.active_query] = id;
-		g_query_state.active_query++;
-		g_query_state.query_issued = true;
-
-		if (g_query_state.mode == GPU_SELECT_NEAREST_SECOND_PASS && g_query_state.index < g_query_state.oldhits) {
-			if (g_query_state.buffer[g_query_state.index * 4 + 3] == id) {
-				g_query_state.index++;
-				return true;
-			}
-			else {
-				return false;
-			}
+		default:  /* ALGO_GL_PICK */
+		{
+			return gpu_select_pick_load_id(id);
 		}
 	}
-
-	return true;
 }
 
 /**
@@ -188,59 +151,27 @@ bool GPU_select_load_id(unsigned int id)
 unsigned int GPU_select_end(void)
 {
 	unsigned int hits = 0;
-	if (!g_query_state.use_gpu_select) {
-		glPopName();
-		hits = glRenderMode(GL_RENDER);
-	}
-	else {
-		int i;
 
-		if (g_query_state.query_issued) {
-			glEndQuery(GL_SAMPLES_PASSED);
+	switch (g_select_state.algorithm) {
+		case ALGO_GL_LEGACY:
+		{
+			glPopName();
+			hits = glRenderMode(GL_RENDER);
+			break;
 		}
-
-		for (i = 0; i < g_query_state.active_query; i++) {
-			unsigned int result;
-			glGetQueryObjectuiv(g_query_state.queries[i], GL_QUERY_RESULT, &result);
-			if (result > 0) {
-				if (g_query_state.mode != GPU_SELECT_NEAREST_SECOND_PASS) {
-					int maxhits = g_query_state.bufsize / 4;
-
-					if (hits < maxhits) {
-						g_query_state.buffer[hits * 4] = 1;
-						g_query_state.buffer[hits * 4 + 1] = 0xFFFF;
-						g_query_state.buffer[hits * 4 + 2] = 0xFFFF;
-						g_query_state.buffer[hits * 4 + 3] = g_query_state.id[i];
-
-						hits++;
-					}
-					else {
-						hits = -1;
-						break;
-					}
-				}
-				else {
-					int j;
-					/* search in buffer and make selected object first */
-					for (j = 0; j < g_query_state.oldhits; j++) {
-						if (g_query_state.buffer[j * 4 + 3] == g_query_state.id[i]) {
-							g_query_state.buffer[j * 4 + 1] = 0;
-							g_query_state.buffer[j * 4 + 2] = 0;
-						}
-					}
-					break;
-				}
-			}
+		case ALGO_GL_QUERY:
+		{
+			hits = gpu_select_query_end();
+			break;
+		}
+		default:  /* ALGO_GL_PICK */
+		{
+			hits = gpu_select_pick_end();
+			break;
 		}
-
-		glDeleteQueries(g_query_state.num_of_queries, g_query_state.queries);
-		MEM_freeN(g_query_state.queries);
-		MEM_freeN(g_query_state.id);
-		glPopAttrib();
-		glColorMask(GL_TRUE, GL_TRUE, GL_TRUE, GL_TRUE);
 	}
 
-	g_query_state.select_is_active = false;
+	g_select_state.select_is_active = false;
 
 	return hits;
 }
@@ -257,3 +188,41 @@ bool GPU_select_query_check_active(void)
 	          GPU_type_matches(GPU_DEVICE_NVIDIA, GPU_OS_UNIX, GPU_DRIVER_OPENSOURCE))));
 
 }
+
+/* ----------------------------------------------------------------------------
+ * Caching
+ *
+ * Support multiple begin/end's as long as they are within the initial region.
+ * Currently only used by ALGO_GL_PICK.
+ */
+
+void GPU_select_cache_begin(void)
+{
+	/* validate on GPU_select_begin, clear if not supported */
+	BLI_assert(g_select_state.use_cache == false);
+	g_select_state.use_cache = true;
+	if (g_select_state.algorithm == ALGO_GL_PICK) {
+		gpu_select_pick_cache_begin();
+	}
+}
+
+void GPU_select_cache_load_id(void)
+{
+	BLI_assert(g_select_state.use_cache == true);
+	if (g_select_state.algorithm == ALGO_GL_PICK) {
+		gpu_select_pick_cache_load_id();
+	}
+}
+
+void GPU_select_cache_end(void)
+{
+	if (g_select_state.algorithm == ALGO_GL_PICK) {
+		gpu_select_pick_cache_end();
+	}
+	g_select_state.use_cache = false;
+}
+
+bool GPU_select_is_cached(void)
+{
+	return g_select_state.use_cache && gpu_select_pick_is_cached();
+}