/**************************************************************************
*
- * Copyright 2007 Tungsten Graphics, Inc., Cedar Park, Texas.
+ * Copyright 2007 VMware, Inc.
* All Rights Reserved.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
* OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT.
- * IN NO EVENT SHALL TUNGSTEN GRAPHICS AND/OR ITS SUPPLIERS BE LIABLE FOR
+ * IN NO EVENT SHALL VMWARE AND/OR ITS SUPPLIERS BE LIABLE FOR
* ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
/*
* Authors:
- * Keith Whitwell <keith@tungstengraphics.com>
+ * Keith Whitwell <keithw@vmware.com>
*/
#include "util/u_cpu_detect.h"
#include "util/u_inlines.h"
#include "util/u_helpers.h"
+#include "util/u_prim.h"
+#include "util/u_format.h"
#include "draw_context.h"
+#include "draw_pipe.h"
+#include "draw_prim_assembler.h"
#include "draw_vs.h"
#include "draw_gs.h"
#include "gallivm/lp_bld_limits.h"
#include "draw_llvm.h"
-static boolean
+boolean
draw_get_option_use_llvm(void)
{
static boolean first = TRUE;
#ifdef PIPE_ARCH_X86
util_cpu_detect();
- /* require SSE2 due to LLVM PR6960. */
+ /* require SSE2 due to LLVM PR6960. XXX Might be fixed by now? */
if (!util_cpu_caps.has_sse2)
value = FALSE;
#endif
if (draw == NULL)
goto err_out;
+ /* we need correct cpu caps for disabling denorms in draw_vbo() */
+ util_cpu_detect();
+
#if HAVE_LLVM
if (try_llvm && draw_get_option_use_llvm()) {
draw->llvm = draw_llvm_create(draw);
if (!draw_init(draw))
goto err_destroy;
+ draw->ia = draw_prim_assembler_create(draw);
+ if (!draw->ia)
+ goto err_destroy;
+
return draw;
err_destroy:
draw->clip_z = TRUE;
draw->pt.user.planes = (float (*) [DRAW_TOTAL_CLIP_PLANES][4]) &(draw->plane[0]);
+ draw->pt.user.eltMax = ~0;
if (!draw_pipeline_init( draw ))
return FALSE;
draw->quads_always_flatshade_last = !draw->pipe->screen->get_param(
draw->pipe->screen, PIPE_CAP_QUADS_FOLLOW_PROVOKING_VERTEX_CONVENTION);
+ draw->floating_point_depth = false;
+
return TRUE;
}
+/*
+ * Called whenever we're starting to draw a new instance.
+ * Some internal structures don't want to have to reset internal
+ * members on each invocation (because their state might have to persist
+ * between multiple primitive restart rendering call) but might have to
+ * for each new instance.
+ * This is particularly the case for primitive id's in geometry shader.
+ */
+void draw_new_instance(struct draw_context *draw)
+{
+ draw_geometry_shader_new_instance(draw->gs.geometry_shader);
+}
+
void draw_destroy( struct draw_context *draw )
{
draw->render->destroy( draw->render );
*/
+ draw_prim_assembler_destroy(draw->ia);
draw_pipeline_destroy( draw );
draw_pt_destroy( draw );
draw_vs_destroy( draw );
/**
- * Specify the Minimum Resolvable Depth factor for polygon offset.
+ * Specify the depth stencil format for the draw pipeline. This function
+ * determines the Minimum Resolvable Depth factor for polygon offset.
* This factor potentially depends on the number of Z buffer bits,
* the rasterization algorithm and the arithmetic performed on Z
- * values between vertex shading and rasterization. It will vary
- * from one driver to another.
+ * values between vertex shading and rasterization.
*/
-void draw_set_mrd(struct draw_context *draw, double mrd)
+void draw_set_zs_format(struct draw_context *draw, enum pipe_format format)
{
- draw->mrd = mrd;
+ const struct util_format_description *desc = util_format_description(format);
+
+ draw->floating_point_depth =
+ (util_get_depth_format_type(desc) == UTIL_FORMAT_TYPE_FLOAT);
+
+ draw->mrd = util_get_depth_format_mrd(desc);
}
draw->rasterizer && draw->rasterizer->depth_clip);
draw->clip_user = draw->rasterizer &&
draw->rasterizer->clip_plane_enable != 0;
+ draw->clip_points_xy = draw->clip_xy &&
+ (!draw->driver.bypass_clip_points ||
+ (draw->rasterizer &&
+ !draw->rasterizer->point_tri_clip));
}
/**
* Some hardware can turn off clipping altogether - in particular any
* hardware with a TNL unit can do its own clipping, even if it is
* relying on the draw module for some other reason.
+ * Setting bypass_clip_points to achieve d3d-style point clipping (the driver
+ * will need to do the "vp scissoring") _requires_ the driver to implement
+ * wide points / point sprites itself (points will still be clipped if rasterizer
+ * point_tri_clip isn't set). Only relevant if bypass_clip_xy isn't set.
*/
void draw_set_driver_clipping( struct draw_context *draw,
boolean bypass_clip_xy,
boolean bypass_clip_z,
- boolean guard_band_xy)
+ boolean guard_band_xy,
+ boolean bypass_clip_points)
{
draw_do_flush( draw, DRAW_FLUSH_STATE_CHANGE );
draw->driver.bypass_clip_xy = bypass_clip_xy;
draw->driver.bypass_clip_z = bypass_clip_z;
draw->driver.guard_band_xy = guard_band_xy;
+ draw->driver.bypass_clip_points = bypass_clip_points;
update_clip_flags(draw);
}
/**
* Set the draw module's viewport state.
*/
-void draw_set_viewport_state( struct draw_context *draw,
- const struct pipe_viewport_state *viewport )
+void draw_set_viewport_states( struct draw_context *draw,
+ unsigned start_slot,
+ unsigned num_viewports,
+ const struct pipe_viewport_state *vps )
{
+ const struct pipe_viewport_state *viewport = vps;
draw_do_flush(draw, DRAW_FLUSH_PARAMETER_CHANGE);
- draw->viewport = *viewport; /* struct copy */
- draw->identity_viewport = (viewport->scale[0] == 1.0f &&
- viewport->scale[1] == 1.0f &&
- viewport->scale[2] == 1.0f &&
- viewport->scale[3] == 1.0f &&
- viewport->translate[0] == 0.0f &&
- viewport->translate[1] == 0.0f &&
- viewport->translate[2] == 0.0f &&
- viewport->translate[3] == 0.0f);
- draw_vs_set_viewport( draw, viewport );
+ debug_assert(start_slot < PIPE_MAX_VIEWPORTS);
+ debug_assert((start_slot + num_viewports) <= PIPE_MAX_VIEWPORTS);
+
+ memcpy(draw->viewports + start_slot, vps,
+ sizeof(struct pipe_viewport_state) * num_viewports);
+
+ draw->identity_viewport = (num_viewports == 1) &&
+ (viewport->scale[0] == 1.0f &&
+ viewport->scale[1] == 1.0f &&
+ viewport->scale[2] == 1.0f &&
+ viewport->scale[3] == 1.0f &&
+ viewport->translate[0] == 0.0f &&
+ viewport->translate[1] == 0.0f &&
+ viewport->translate[2] == 0.0f &&
+ viewport->translate[3] == 0.0f);
}
*/
void
draw_set_mapped_vertex_buffer(struct draw_context *draw,
- unsigned attr, const void *buffer)
+ unsigned attr, const void *buffer,
+ size_t size)
{
- draw->pt.user.vbuffer[attr] = buffer;
+ draw->pt.user.vbuffer[attr].map = buffer;
+ draw->pt.user.vbuffer[attr].size = size;
}
uint n;
slot = draw_find_shader_output(draw, semantic_name, semantic_index);
- if (slot > 0) {
+ if (slot >= 0) {
return slot;
}
}
}
+/**
+ * Prepare outputs slots from the draw module
+ *
+ * Certain parts of the draw module can emit additional
+ * outputs that can be quite useful to the backends, a good
+ * example of it is the process of decomposing primitives
+ * into wireframes (aka. lines) which normally would lose
+ * the face-side information, but using this method we can
+ * inject another shader output which passes the original
+ * face side information to the backend.
+ */
+void
+draw_prepare_shader_outputs(struct draw_context *draw)
+{
+ draw_remove_extra_vertex_attribs(draw);
+ draw_prim_assembler_prepare_outputs(draw->ia);
+ draw_unfilled_prepare_outputs(draw, draw->pipeline.unfilled);
+ if (draw->pipeline.aapoint)
+ draw_aapoint_prepare_outputs(draw, draw->pipeline.aapoint);
+ if (draw->pipeline.aaline)
+ draw_aaline_prepare_outputs(draw, draw->pipeline.aaline);
+}
/**
* Ask the draw module for the location/slot of the given vertex attribute in
* attributes (such as texcoords for AA lines). The driver can call this
* function to find those attributes.
*
- * Zero is returned if the attribute is not found since this is
- * a don't care / undefined situtation. Returning -1 would be a bit more
- * work for the drivers.
+ * -1 is returned if the attribute is not found since this is
+ * an undefined situation. Note, that zero is valid and can
+ * be used by any of the attributes, because position is not
+ * required to be attribute 0 or even at all present.
*/
int
draw_find_shader_output(const struct draw_context *draw,
}
}
- return 0;
+ return -1;
}
}
+/**
+ * Return total number of the vertex shader outputs. This function
+ * also counts any extra vertex output attributes that may
+ * be filled in by some draw stages (such as AA point, AA line,
+ * front face).
+ */
+uint
+draw_total_vs_outputs(const struct draw_context *draw)
+{
+ const struct tgsi_shader_info *info = &draw->vs.vertex_shader->info;
+
+ return info->num_outputs + draw->extra_shader_outputs.num;;
+}
+
+/**
+ * Return total number of the geometry shader outputs. This function
+ * also counts any extra geometry output attributes that may
+ * be filled in by some draw stages (such as AA point, AA line, front
+ * face).
+ */
+uint
+draw_total_gs_outputs(const struct draw_context *draw)
+{
+ const struct tgsi_shader_info *info;
+
+ if (!draw->gs.geometry_shader)
+ return 0;
+
+ info = &draw->gs.geometry_shader->info;
+
+ return info->num_outputs + draw->extra_shader_outputs.num;
+}
+
+
/**
* Provide TGSI sampler objects for vertex/geometry shaders that use
* texture fetches. This state only needs to be set once per context.
* This might only be used by software drivers for the time being.
*/
void
-draw_texture_samplers(struct draw_context *draw,
- uint shader,
- uint num_samplers,
- struct tgsi_sampler **samplers)
+draw_texture_sampler(struct draw_context *draw,
+ uint shader,
+ struct tgsi_sampler *sampler)
{
if (shader == PIPE_SHADER_VERTEX) {
- draw->vs.tgsi.num_samplers = num_samplers;
- draw->vs.tgsi.samplers = samplers;
+ draw->vs.tgsi.sampler = sampler;
} else {
debug_assert(shader == PIPE_SHADER_GEOMETRY);
- draw->gs.tgsi.num_samplers = num_samplers;
- draw->gs.tgsi.samplers = samplers;
+ draw->gs.tgsi.sampler = sampler;
}
}
*/
void
draw_set_indexes(struct draw_context *draw,
- const void *elements, unsigned elem_size)
+ const void *elements, unsigned elem_size,
+ unsigned elem_buffer_space)
{
assert(elem_size == 0 ||
elem_size == 1 ||
elem_size == 4);
draw->pt.user.elts = elements;
draw->pt.user.eltSizeIB = elem_size;
+ if (elem_size)
+ draw->pt.user.eltMax = elem_buffer_space / elem_size;
+ else
+ draw->pt.user.eltMax = 0;
}
}
+/**
+ * Return the index of the shader output which will contain the
+ * viewport index.
+ */
+uint
+draw_current_shader_viewport_index_output(const struct draw_context *draw)
+{
+ if (draw->gs.geometry_shader)
+ return draw->gs.geometry_shader->viewport_index_output;
+ return 0;
+}
+
+/**
+ * Returns true if there's a geometry shader bound and the geometry
+ * shader writes out a viewport index.
+ */
+boolean
+draw_current_shader_uses_viewport_index(const struct draw_context *draw)
+{
+ if (draw->gs.geometry_shader)
+ return draw->gs.geometry_shader->info.writes_viewport_index;
+ return FALSE;
+}
+
+
/**
* Return the index of the shader output which will contain the
* vertex position.
uint
draw_current_shader_clipdistance_output(const struct draw_context *draw, int index)
{
+ debug_assert(index < PIPE_MAX_CLIP_OR_CULL_DISTANCE_ELEMENT_COUNT);
+ if (draw->gs.geometry_shader)
+ return draw->gs.geometry_shader->clipdistance_output[index];
return draw->vs.clipdistance_output[index];
}
+
+uint
+draw_current_shader_num_written_clipdistances(const struct draw_context *draw)
+{
+ if (draw->gs.geometry_shader)
+ return draw->gs.geometry_shader->info.num_written_clipdistance;
+ return draw->vs.vertex_shader->info.num_written_clipdistance;
+}
+
+
+uint
+draw_current_shader_culldistance_output(const struct draw_context *draw, int index)
+{
+ debug_assert(index < PIPE_MAX_CLIP_OR_CULL_DISTANCE_ELEMENT_COUNT);
+ if (draw->gs.geometry_shader)
+ return draw->gs.geometry_shader->culldistance_output[index];
+ return draw->vs.vertex_shader->culldistance_output[index];
+}
+
+uint
+draw_current_shader_num_written_culldistances(const struct draw_context *draw)
+{
+ if (draw->gs.geometry_shader)
+ return draw->gs.geometry_shader->info.num_written_culldistance;
+ return draw->vs.vertex_shader->info.num_written_culldistance;
+}
+
/**
* Return a pointer/handle for a driver/CSO rasterizer object which
* disabled culling, stippling, unfilled tris, etc.
rast.scissor = scissor;
rast.flatshade = flatshade;
rast.front_ccw = 1;
- rast.gl_rasterization_rules = draw->rasterizer->gl_rasterization_rules;
+ rast.half_pixel_center = draw->rasterizer->half_pixel_center;
+ rast.bottom_edge_rule = draw->rasterizer->bottom_edge_rule;
+ rast.clip_halfz = draw->rasterizer->clip_halfz;
draw->rasterizer_no_cull[scissor][flatshade] =
pipe->create_rasterizer_state(pipe, &rast);
draw->so.num_targets = num_targets;
}
-void
-draw_set_mapped_so_buffers(struct draw_context *draw,
- void *buffers[PIPE_MAX_SO_BUFFERS],
- unsigned num_buffers)
-{
-}
-
-void
-draw_set_so_state(struct draw_context *draw,
- struct pipe_stream_output_info *state)
-{
- memcpy(&draw->so.state,
- state,
- sizeof(struct pipe_stream_output_info));
-}
-
void
draw_set_sampler_views(struct draw_context *draw,
unsigned shader_stage,
draw->num_samplers[shader_stage] = num;
#ifdef HAVE_LLVM
- if (draw->llvm && shader_stage == PIPE_SHADER_VERTEX)
- draw_llvm_set_sampler_state(draw);
+ if (draw->llvm)
+ draw_llvm_set_sampler_state(draw, shader_stage);
#endif
}
uint32_t img_stride[PIPE_MAX_TEXTURE_LEVELS],
uint32_t mip_offsets[PIPE_MAX_TEXTURE_LEVELS])
{
- if (shader_stage == PIPE_SHADER_VERTEX) {
#ifdef HAVE_LLVM
- if (draw->llvm)
- draw_llvm_set_mapped_texture(draw,
- sview_idx,
- width, height, depth, first_level,
- last_level, base_ptr,
- row_stride, img_stride, mip_offsets);
+ if (draw->llvm)
+ draw_llvm_set_mapped_texture(draw,
+ shader_stage,
+ sview_idx,
+ width, height, depth, first_level,
+ last_level, base_ptr,
+ row_stride, img_stride, mip_offsets);
#endif
- }
}
/**
return draw_get_shader_param_no_llvm(shader, param);
}
+/**
+ * Enables or disables collection of statistics.
+ *
+ * Draw module is capable of generating statistics for the vertex
+ * processing pipeline. Collection of that data isn't free and so
+ * it's disabled by default. The users of the module can enable
+ * (or disable) this functionality through this function.
+ * The actual data will be emitted through the VBUF interface,
+ * the 'pipeline_statistics' callback to be exact.
+ */
+void
+draw_collect_pipeline_statistics(struct draw_context *draw,
+ boolean enable)
+{
+ draw->collect_statistics = enable;
+}
+
+/**
+ * Computes clipper invocation statistics.
+ *
+ * Figures out how many primitives would have been
+ * sent to the clipper given the specified
+ * prim info data.
+ */
+void
+draw_stats_clipper_primitives(struct draw_context *draw,
+ const struct draw_prim_info *prim_info)
+{
+ if (draw->collect_statistics) {
+ unsigned start, i;
+ for (start = i = 0;
+ i < prim_info->primitive_count;
+ start += prim_info->primitive_lengths[i], i++)
+ {
+ draw->statistics.c_invocations +=
+ u_decomposed_prims_for_vertices(prim_info->prim,
+ prim_info->primitive_lengths[i]);
+ }
+ }
+}
+
+
+/**
+ * Returns true if the draw module will inject the frontface
+ * info into the outputs.
+ *
+ * Given the specified primitive and rasterizer state
+ * the function will figure out if the draw module
+ * will inject the front-face information into shader
+ * outputs. This is done to preserve the front-facing
+ * info when decomposing primitives into wireframes.
+ */
+boolean
+draw_will_inject_frontface(const struct draw_context *draw)
+{
+ unsigned reduced_prim = u_reduced_prim(draw->pt.prim);
+ const struct pipe_rasterizer_state *rast = draw->rasterizer;
+
+ if (reduced_prim != PIPE_PRIM_TRIANGLES) {
+ return FALSE;
+ }
+
+ return (rast &&
+ (rast->fill_front != PIPE_POLYGON_MODE_FILL ||
+ rast->fill_back != PIPE_POLYGON_MODE_FILL));
+}