Sign in to follow this  

DX12 My Software Triangle Rasterizer

This topic is 3843 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Hi all! Last weekend I took the time to code a class which can rasterize triangles basing my work on some code I could find on the net, particularly this one over at devmaster. Over the course of this week I extended it and now I would say it is a complete solution. Features:
  • The rasterizer can support an arbitrary number of render targets (you will most likely use two, a color buffer and a depth buffer)
  • The rasterization is completely decoupled from the actual shading of the visible pixels. You can configure the rasterizer with a pixel shader which does the actual work of computing and assigning a color value.
  • It does only use integer math because i intend to use it on the GP2X which does not have an FPU.
  • The rasterizer is tile based. Currently it uses blocks of 8x8 pixels.
  • It interpolates an arbitrary number of integer varyings across the triangle (so you can use fixed point here). This is done perspectively correct for the corners of each 8x8 block and affine within each block to avoid the costly per pixel divide.
  • It supports a clipping rectangle.
  • It provides a means for the pixel shader to compute the derivative of the interpolated varyings. This is needed for example to compute the texture mimmap level from the texture coordinates.
  • It allows for an early depth test. For example the shader could store the minimum depth value for each 8x8 block and than discad a whole block if the minimum expected depth value for this block is greater than the one stored.
  • The source code is actually quite short ~600 lines with a lot of comments.
The only problem I can see right now is with small or large but thin triangles. Because the rasterizer is tile based it must at least scan a whole 8x8 block and test each pixel if it is inside the triangle or not. Large triangles are handled quite efficiently since for the inner part only the corners are tested for inout. What do you think about this. How big a performance problem might this be when targeting the GP2X? I include the code here for everyoune to look at. I would be very thankful for any input of possible improvements.
Header file:
/*
Copyright (c) 2007, Markus Trenkwalder

All rights reserved.

Redistribution and use in source and binary forms, with or without 
modification, are permitted provided that the following conditions are met:

* Redistributions of source code must retain the above copyright notice, 
  this list of conditions and the following disclaimer.

* Redistributions in binary form must reproduce the above copyright notice,
  this list of conditions and the following disclaimer in the documentation 
  and/or other materials provided with the distribution.

* Neither the name of the <ORGANIZATION> nor the names of its contributors 
  may be used to endorse or promote products derived from this software 
  without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/

#ifndef RASTERIZER_EE4F8987_BDA2_434d_A01F_BB3446E535C3
#define RASTERIZER_EE4F8987_BDA2_434d_A01F_BB3446E535C3

class Rasterizer {
public:
	// some constants
	static const int MAX_RENDER_TARGETS = 2;
	static const int MAX_VARYING = 8;
	static const int BLOCK_SIZE = 8;	

public:
	// Type definitions
	struct Vertex {
		int x, y; // in 28.4 fixed point
		int z; // range from 0 to 0x7fffffff
		int w; // in 16.16 fixed point
		int varyings[MAX_VARYING];
	};

	// holds the pointers to each render target. 
	// These will be passed to the fragment shader which then can write to
	// the pointed to location. Note: Only the first n pointers will be
	// valid where n is the current number of render targets
	struct BufferPointers {
		void* ptr[MAX_RENDER_TARGETS];
	};

	// This is the data the fragment shader gets
	struct FragmentData {
		int z;
		int varying[MAX_VARYING];
	};

	typedef FragmentData PixelBlock[BLOCK_SIZE][BLOCK_SIZE] ;

	class RenderTarget {
	public:
		virtual int width() = 0;
		virtual int height() = 0;
		virtual int stride() = 0;
		virtual void *buffer_pointer() = 0;
		virtual int element_size() = 0;
		virtual void clear(int x, int y, int w, int h) = 0;
	};

	class FragmentShader {
	public:
		// This provides a means for an early depth test.
		// x and y are the coordinates of the upper left corner of the current block.
		// If the shader somewhere stores the minimum z of each block that value
		// can be compared to the parameter z.
		// returns false when the depth test failed. In this case the whole block 
		// can be culled.
		virtual bool early_depth_test(int x, int y, int z) { return true; }

		// This notifies the shader of any render target clears.
		// This is meant to be used in conjunction with the early depth test to update
		// any buffers used
		virtual void clear(int target, int x, int y, int w, int h) {}

		// To compute the mipmap level of detail one needs the derivativs in x and y of 
		// the texture coordinates. These can be computed from the values in the pixel
		// block since all the fragment values have alredy been computed for this block
		// when this is called
		virtual void prepare_for_block(int x, int y, PixelBlock b) {}

		// This tells the rasterizer how many varyings this fragment shader needs
		virtual int  varying_count() = 0;

		// This is called once for each visible fragment inside the triangle
		// x and y are the coordinates within the block [0, BLOCK_SIZE[
		// the pixel block is indexed with p[y][x] !!!
		virtual void shade(const BufferPointers&, const PixelBlock& b, int x, int y) = 0;
	};

private:
	// Variables
	struct RenderTargetParams {
		int count;
		int minwidth, minheight;

		// cache these params to avoid too 
		// many virtual function calls
		int stride[MAX_RENDER_TARGETS];
		int element_size[MAX_RENDER_TARGETS];
	} rendertarget_params_;

	RenderTarget *rendertargets_[MAX_RENDER_TARGETS];
	FragmentShader *fragment_shader_;

	struct {
		int x0, y0, x1, y1;
	} clip_rect_;

private:
	bool setup_valid();

public:
	// constructor
	Rasterizer();

public:
	// main interface

	// Set the render targets.
	// This resets the clipping rectangle
	void rendertargets(int n, RenderTarget* rt[]);

	// set the fragment shader
	void fragment_shader(FragmentShader *fs);

	void clear();
	void clear(int target);

	void clip_rect(int x, int y, int w, int h);

	// The triangle must be counter clockwise in screen space in order to be
	// drawn.
	void draw_triangle(const Vertex &v1, const Vertex &v2, const Vertex &v3);
};

#endif




Implementation file:
/*
Copyright (c) 2007, Markus Trenkwalder

All rights reserved.

Redistribution and use in source and binary forms, with or without 
modification, are permitted provided that the following conditions are met:

* Redistributions of source code must retain the above copyright notice, 
  this list of conditions and the following disclaimer.

* Redistributions in binary form must reproduce the above copyright notice,
  this list of conditions and the following disclaimer in the documentation 
  and/or other materials provided with the distribution.

* Neither the name of the <ORGANIZATION> nor the names of its contributors 
  may be used to endorse or promote products derived from this software 
  without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/

#include "rasterizer.h"

#include <cmath>
#include <cassert>
#include <algorithm>

#ifndef _MSC_VER
#include <stdint.h>
#else
#include "stdint.h"
#endif

////////////////////////////////////////////////////////////////////////////////
// utility functions

namespace {
	inline int min(int a, int b, int c)
	{
		return std::min(std::min(a,b), c);
	}

	inline int max(int a, int b, int c) 
	{
		return std::max(std::max(a,b), c);
	}

	inline void compute_plane(
		int v0x, int v0y,
		int v1x, int v1y,
		int v2x, int v2y,
		int z0, int z1, int z2, int64_t plane[4])
	{
	   const int px = v1x - v0x;
	   const int py = v1y - v0y;
	   const int pz = z1 - z0;

	   const int qx = v2x - v0x;
	   const int qy = v2y - v0y;
	   const int qz = z2 - z0;

	   /* Crossproduct "(a,b,c):= dv1 x dv2" is orthogonal to plane. */
	   const int64_t a = (int64_t)py * qz - (int64_t)pz * qy;
	   const int64_t b = (int64_t)pz * qx - (int64_t)px * qz;
	   const int64_t c = (int64_t)px * qy - (int64_t)py * qx;
	   /* Point on the plane = "r*(a,b,c) + w", with fixed "r" depending
		  on the distance of plane from origin and arbitrary "w" parallel
		  to the plane. */
	   /* The scalar product "(r*(a,b,c)+w)*(a,b,c)" is "r*(a^2+b^2+c^2)",
		  which is equal to "-d" below. */
	   const int64_t d = -(a * v0x + b * v0y + c * z0);

	   plane[0] = a;
	   plane[1] = b;
	   plane[2] = c;
	   plane[3] = d;
	}

	inline int solve_plane(int x, int y, const int64_t plane[4])
	{
	   assert(plane[2] != 0);
	   return (int)((plane[3] + plane[0] * x + plane[1] * y) / -plane[2]);
	}

	template <int denominator>
	inline void floor_divmod(int numerator, int &floor, int &mod)
	{
		assert(denominator > 0);
		if(numerator >= 0) {
			// positive case, C is okay
			floor = numerator / denominator;
			mod = numerator % denominator;
		} else {
			// Numerator is negative, do the right thing
			floor = -((-numerator) / denominator);
			mod = (-numerator) % denominator;
			if(mod) {
				// there is a remainder
				floor--; mod = denominator - mod;
			}
		}
	}

	// Fixed point division
	template <int p>
	inline int32_t fixdiv(int32_t a, int32_t b)
	{
	#if 0
		return (int32_t)((((int64_t)a) << p) / b);
	#else	
		// The following produces the same results as the above but gcc 4.0.3 
		// generates fewer instructions (at least on the ARM processor).
		union {
			int64_t a;
			struct {
				int32_t l;
				int32_t h;
			};
		} x;
		
		x.l = a << p;
		x.h = a >> (sizeof(int32_t) * 8 - p);
		return (int32_t)(x.a / b);
	#endif
	}

	// Perform a fixed point multiplication using a 64-bit intermediate result to
	// prevent overflow problems.
	template <int p>
	inline int32_t fixmul(int32_t a, int32_t b)
	{
		return (int32_t)(((int64_t)a * b) >> p);
	}
} // end anonymous namespace

////////////////////////////////////////////////////////////////////////////////

Rasterizer::Rasterizer() :
	fragment_shader_(0)	
{
	rendertarget_params_.count = 0;
}

bool Rasterizer::setup_valid()
{
	return rendertarget_params_.count >= 1 &&
		fragment_shader_ != 0;
}

void Rasterizer::clear()
{
	for (int i = 0; i < rendertarget_params_.count; ++i) 
		clear(i);
}

void Rasterizer::clear(int target)
{
	assert(target <= rendertarget_params_.count);
	rendertargets_[target]->clear(0, 0, 
		rendertarget_params_.minwidth, rendertarget_params_.minheight);

	// notify shader about clear (might want to update internal data structutes)
	if (fragment_shader_)
		fragment_shader_->clear(target, 0, 0, 
		rendertarget_params_.minwidth, rendertarget_params_.minheight);
}

void Rasterizer::clip_rect(int x, int y, int w, int h)
{
	if (rendertarget_params_.count == 0) return;

	clip_rect_.x0 = std::max(0, x);
	clip_rect_.y0 = std::max(0, y);
	clip_rect_.x1 = std::min(x + w, rendertarget_params_.minwidth);
	clip_rect_.y1 = std::min(y + h, rendertarget_params_.minheight);
}

////////////////////////////////////////////////////////////////////////////////
// main interface

void Rasterizer::rendertargets(int n, RenderTarget* rt[])
{
	assert(n <= MAX_RENDER_TARGETS);
	RenderTargetParams &rtp = rendertarget_params_;
	rtp.count = n;

	if (n == 0) return;

	rtp.minwidth = rt[0]->width();
	rtp.minheight = rt[0]->height();
	for (int i = 0; i < n; ++i) {
		rendertargets_[i] = rt[i];
		rtp.minwidth = std::min(rtp.minwidth, rt[i]->width());
		rtp.minheight = std::min(rtp.minheight, rt[i]->height());

		// cache these to avoid too many virtual function calls later
		rtp.element_size[i] = rt[i]->element_size();
		rtp.stride[i] = rt[i]->stride();
	}

	clip_rect_.x0 = 0;
	clip_rect_.y0 = 0;
	clip_rect_.x1 = rtp.minwidth;
	clip_rect_.y1 = rtp.minheight;
}

void Rasterizer::fragment_shader(FragmentShader *fs)
{
	assert(fs != 0);
	fragment_shader_ = fs;
}

void Rasterizer::draw_triangle(const Vertex &v1, const Vertex &v2, const Vertex &v3)
{
	if (!setup_valid()) return;

	int64_t zPlane[4];
	int64_t wPlane[4];
	int64_t vPlane[MAX_VARYING][4];
	compute_plane(v1.x, v1.y, v2.x, v2.y, v3.x, v3.y, v1.z, v2.z, v3.z, zPlane);
	compute_plane(v1.x, v1.y, v2.x, v2.y, v3.x, v3.y, // interpolate 1/w across triangle
		fixdiv<16>(1 << 16, v1.w),
		fixdiv<16>(1 << 16, v2.w),
		fixdiv<16>(1 << 16, v3.w),
		wPlane);
	int varying_count = fragment_shader_->varying_count();
	for (int i = 0; i < varying_count; ++i)
		compute_plane(
			v1.x, v1.y, v2.x, v2.y, v3.x, v3.y, 
			fixdiv<16>(v1.varyings[i], v1.w), 
			fixdiv<16>(v2.varyings[i], v2.w),
			fixdiv<16>(v3.varyings[i], v3.w),
			vPlane[i]
		);

	// Deltas
	const int DX12 = v1.x - v2.x;
	const int DX23 = v2.x - v3.x;
	const int DX31 = v3.x - v1.x;

	const int DY12 = v1.y - v2.y;
	const int DY23 = v2.y - v3.y;
	const int DY31 = v3.y - v1.y;

	// Fixed-point deltas
	const int FDX12 = DX12 << 4;
	const int FDX23 = DX23 << 4;
	const int FDX31 = DX31 << 4;

	const int FDY12 = DY12 << 4;
	const int FDY23 = DY23 << 4;
	const int FDY31 = DY31 << 4;

	// Bounding rectangle
	int minx = (min(v1.x, v2.x, v3.x) + 0xF) >> 4;
	int maxx = (max(v1.x, v2.x, v3.x) + 0xF) >> 4;
	int miny = (min(v1.y, v2.y, v3.y) + 0xF) >> 4;
	int maxy = (max(v1.y, v2.y, v3.y) + 0xF) >> 4;

	// consider clipping rectangle
	minx = std::max(minx, clip_rect_.x0);
	maxx = std::min(maxx, clip_rect_.x1);
	miny = std::max(miny, clip_rect_.y0);
	maxy = std::min(maxy, clip_rect_.y1);

	// Start in corner of 8x8 block
	minx &= ~(BLOCK_SIZE - 1);
	miny &= ~(BLOCK_SIZE - 1);

	BufferPointers buffers;

	for (int i = 0; i < rendertarget_params_.count; ++i)
		buffers.ptr[i] = (char*)rendertargets_[i]->buffer_pointer() + miny * rendertargets_[i]->stride();
	
	// Half-edge constants
	int C1 = DY12 * v1.x - DX12 * v1.y;
	int C2 = DY23 * v2.x - DX23 * v2.y;
	int C3 = DY31 * v3.x - DX31 * v3.y;

	// Correct for fill convention
	if(DY12 < 0 || (DY12 == 0 && DX12 > 0)) C1++;
	if(DY23 < 0 || (DY23 == 0 && DX23 > 0)) C2++;
	if(DY31 < 0 || (DY31 == 0 && DX31 > 0)) C3++;

	// Loop through blocks
	for(int y = miny; y < maxy; y += BLOCK_SIZE)
	{
		for(int x = minx; x < maxx; x += BLOCK_SIZE)
		{
			// Corners of block
			int x0 = x << 4;
			int x1 = (x + BLOCK_SIZE - 1) << 4;
			int y0 = y << 4;
			int y1 = (y + BLOCK_SIZE - 1) << 4;

			// Evaluate half-space functions
			bool a00 = C1 + DX12 * y0 - DY12 * x0 > 0;
			bool a10 = C1 + DX12 * y0 - DY12 * x1 > 0;
			bool a01 = C1 + DX12 * y1 - DY12 * x0 > 0;
			bool a11 = C1 + DX12 * y1 - DY12 * x1 > 0;
			int a = (a00 << 0) | (a10 << 1) | (a01 << 2) | (a11 << 3);
    
			bool b00 = C2 + DX23 * y0 - DY23 * x0 > 0;
			bool b10 = C2 + DX23 * y0 - DY23 * x1 > 0;
			bool b01 = C2 + DX23 * y1 - DY23 * x0 > 0;
			bool b11 = C2 + DX23 * y1 - DY23 * x1 > 0;
			int b = (b00 << 0) | (b10 << 1) | (b01 << 2) | (b11 << 3);
    
			bool c00 = C3 + DX31 * y0 - DY31 * x0 > 0;
			bool c10 = C3 + DX31 * y0 - DY31 * x1 > 0;
			bool c01 = C3 + DX31 * y1 - DY31 * x0 > 0;
			bool c11 = C3 + DX31 * y1 - DY31 * x1 > 0;
			int c = (c00 << 0) | (c10 << 1) | (c01 << 2) | (c11 << 3);

			// Skip block when outside an edge
			if(a == 0x0 || b == 0x0 || c == 0x0) continue;

			#define CLIP_TEST(X, Y) 				((X) >= clip_rect_.x0 && (X) < clip_rect_.x1 && 				(Y) >= clip_rect_.y0 && (Y) < clip_rect_.y1)

			// test for the clipping rectangle
			bool clip00 = CLIP_TEST(x, y);
			bool clip10 = CLIP_TEST(x + 7, y);
			bool clip01 = CLIP_TEST(x, y + 7);
			bool clip11 = CLIP_TEST(x + 7, y + 7);

			// skip block if all is clippled
			if (!clip00 && !clip10 && !clip01 && !clip11) continue;
			bool clip_all_in = clip00 && clip10 && clip01 && clip11;
			
			//! compute attribute interpolants at corners
			FragmentData f00; 
			FragmentData f10; 
			FragmentData f01; 
			FragmentData f11; 

			int xx1 = (x + BLOCK_SIZE) << 4;
			int yy1 = (y + BLOCK_SIZE) << 4;
			f00.z = solve_plane(x0, y0, zPlane);			
			f10.z = solve_plane(xx1, y0, zPlane);
			f01.z = solve_plane(x0, yy1, zPlane);
			f11.z = solve_plane(xx1, yy1, zPlane);
			
			if (!fragment_shader_->early_depth_test(x, y, std::min(std::min(std::min(f00.z, f10.z), f01.z), f11.z)))
				continue;

			int w00 = fixdiv<16>(1 << 16, solve_plane(x0, y0, wPlane));
			int w10 = fixdiv<16>(1 << 16, solve_plane(xx1, y0, wPlane));
			int w01 = fixdiv<16>(1 << 16, solve_plane(x0, yy1, wPlane));
			int w11 = fixdiv<16>(1 << 16, solve_plane(xx1, yy1, wPlane));
			for (int i = 0; i < varying_count; ++i) {
				f00.varying[i] = fixmul<16>(solve_plane(x0, y0, vPlane[i]), w00);
				f10.varying[i] = fixmul<16>(solve_plane(xx1, y0, vPlane[i]), w10);
				f01.varying[i] = fixmul<16>(solve_plane(x0, yy1, vPlane[i]), w01);
				f11.varying[i] = fixmul<16>(solve_plane(xx1, yy1, vPlane[i]), w11);
			}

			//! compute attribute step y left and right
			struct varying_step_t {
				struct step_info_t {
					int step;
					int rem;
					int error_term;
					step_info_t():error_term(0){}

					int dostep() {
						int r = step;
						error_term += rem;
						if (error_term >= BLOCK_SIZE) { 
							error_term -= BLOCK_SIZE;
							r++;
						}
						return r;
					}
				};

				step_info_t z;
				step_info_t varying[MAX_VARYING];

				varying_step_t(FragmentData& p1, FragmentData& p2, int vc)
				{
					floor_divmod<BLOCK_SIZE>(p2.z - p1.z, z.step, z.rem);
					for (int i = 0; i < vc; ++i) {
						floor_divmod<BLOCK_SIZE>(p2.varying[i] - p1.varying[i], varying[i].step, varying[i].rem);
					}
				}
			};
			
			varying_step_t step_left(f00, f01, varying_count);
			varying_step_t step_right(f10, f11, varying_count);			

			BufferPointers block_buffers = buffers;

			#define RENDER_TARGET_LOOP 				for (int i = 0; i < rendertarget_params_.count; ++i)

			#define STEP_POINTERS_BY_ELEMENTSIZE(VAR, FACTOR) 				{ 					RENDER_TARGET_LOOP 					(char*&)VAR.ptr[i] += FACTOR * rendertarget_params_.element_size[i]; 				}
				
			#define	STEP_POINTERS_BY_STRIDE(VAR) 				{ 					RENDER_TARGET_LOOP 					(char*&)VAR.ptr[i] += rendertarget_params_.stride[i]; 				}

			#define STEP_FRAGMENTDATA(FDVAR, STEPVAR) 				{ 					FDVAR.z += STEPVAR.z.dostep(); 					for (int i = 0; i < varying_count; ++i) 						FDVAR.varying[i] += STEPVAR.varying[i].dostep(); 				}
			
			// only copy the neccessary varyings
			#define EFFICIENT_COPY(SRC, DST) 				{ 					DST.z = SRC.z; 					for (int i = 0; i < varying_count; ++i) 						DST.varying[i] = SRC.varying[i]; 				}
			
			#define BLOCK_BEGIN 				fragment_shader_->prepare_for_block(x, y, pixel_block); 				for (int iy = 0; iy < BLOCK_SIZE; ++iy) { 					BufferPointers inner = block_buffers; 					STEP_POINTERS_BY_ELEMENTSIZE(inner, x); 					for (int ix = 0; ix < BLOCK_SIZE; ++ix) {

			#define BLOCK_END 						STEP_POINTERS_BY_ELEMENTSIZE(inner, 1); 					} 					STEP_POINTERS_BY_STRIDE(block_buffers); 				}

			PixelBlock pixel_block;
			
			bool skip_flag[BLOCK_SIZE][BLOCK_SIZE];
			memset(skip_flag, 0, sizeof(skip_flag));

			if (!clip_all_in) {
				for (int iy = 0; iy < BLOCK_SIZE; ++iy)
				for (int ix = 0; ix < BLOCK_SIZE; ++ix)
					if (!CLIP_TEST(ix + x, iy + y))
						skip_flag[iy][ix] = true;
			}

			// Accept whole block when totally covered
			if(a == 0xF && b == 0xF && c == 0xF)
			{
				// first compute all fragment data
				for(int iy = 0; iy < BLOCK_SIZE; iy++)
				{
					//! compute attribute step x for this scanline
					varying_step_t stepx(f00, f10, varying_count);
					FragmentData fragment_data = f00;

					for(int ix = 0; ix < BLOCK_SIZE; ix++)
					{
						EFFICIENT_COPY(fragment_data, pixel_block[iy][ix]);
						STEP_FRAGMENTDATA(fragment_data, stepx);
					}

					//! step left and right attrib y
					STEP_FRAGMENTDATA(f00, step_left);
					STEP_FRAGMENTDATA(f10, step_right);
				}

				//! fragment_shader_block (can now use derivatives of attributes)
				if (clip_all_in) {
					BLOCK_BEGIN
						fragment_shader_->shade(inner, pixel_block, ix, iy);
					BLOCK_END
				} else {
					BLOCK_BEGIN
						if (!skip_flag[iy][ix]) 
							fragment_shader_->shade(inner, pixel_block, ix, iy);
					BLOCK_END
				}
			}
			else // Partially covered block
			{
				int CY1 = C1 + DX12 * y0 - DY12 * x0;
				int CY2 = C2 + DX23 * y0 - DY23 * x0;
				int CY3 = C3 + DX31 * y0 - DY31 * x0;

				for(int iy = 0; iy < BLOCK_SIZE; iy++)
				{
					int CX1 = CY1;
					int CX2 = CY2;
					int CX3 = CY3;
					
					//! compute attribute step x for this scanline
					varying_step_t stepx(f00, f10, varying_count);
					
					FragmentData fragment_data = f00;

					for(int ix = 0; ix < BLOCK_SIZE; ix++)
					{
						if(!(CX1 > 0 && CX2 > 0 && CX3 > 0))
							skip_flag[iy][ix] = true;

						// we still need to do this since the fragment shader might want
						// to compute the derivative of attibutes
						EFFICIENT_COPY(fragment_data, pixel_block[iy][ix]);

						CX1 -= FDY12;
						CX2 -= FDY23;
						CX3 -= FDY31;

						STEP_FRAGMENTDATA(fragment_data, stepx);
					}
					
					CY1 += FDX12;
					CY2 += FDX23;
					CY3 += FDX31;

					//! step left and right attrib y
					STEP_FRAGMENTDATA(f00, step_left);
					STEP_FRAGMENTDATA(f10, step_right);
				}
				
				//! fragment_shader_block (can now use derivatives of attributes)
				BLOCK_BEGIN
					if (!skip_flag[iy][ix])
						fragment_shader_->shade(inner, pixel_block, ix, iy);
				BLOCK_END
			}
		}

		for (int i = 0; i < rendertarget_params_.count; ++i)
			(char*&)buffers.ptr[i] += BLOCK_SIZE * rendertargets_[i]->stride();

	}
}




Test program:
/*
Copyright (c) 2007, Markus Trenkwalder

All rights reserved.

Redistribution and use in source and binary forms, with or without 
modification, are permitted provided that the following conditions are met:

* Redistributions of source code must retain the above copyright notice, 
  this list of conditions and the following disclaimer.

* Redistributions in binary form must reproduce the above copyright notice,
  this list of conditions and the following disclaimer in the documentation 
  and/or other materials provided with the distribution.

* Neither the name of the <ORGANIZATION> nor the names of its contributors 
  may be used to endorse or promote products derived from this software 
  without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/

#include "SDL.h"
#include "rasterizer.h"

#include <cmath>
#include <algorithm>

class SDL_SurfaceRenderTarget : public Rasterizer::RenderTarget {
	SDL_Surface *surface;
public:
	SDL_SurfaceRenderTarget(SDL_Surface *s):surface(s) {}
	virtual int width() { return surface->w; }
	virtual int height() { return surface->h; }
	virtual int stride() { return surface->pitch; }
	virtual int element_size() { return sizeof(int); }
	virtual void* buffer_pointer() { return surface->pixels; }
	virtual void clear(int x, int y, int w, int h) {}
};

class TestFragmentShader : public Rasterizer::FragmentShader {
public:
	virtual int varying_count() { return 3; }
	virtual void shade(const Rasterizer::BufferPointers& ptrs, const Rasterizer::PixelBlock& block, int x, int y)
	{
		unsigned int* color_buffer = (unsigned int*)ptrs.ptr[0];

		// unfortunaltely at the corners of the triangle we can get negative
		// values for the interpolants -> std::max
		int r = std::max(0, block[y][x].varying[0]);
		int g = std::max(0, block[y][x].varying[1]);
		int b = std::max(0, block[y][x].varying[2]);
		int color = r << 16 | g << 8 | b;
		*color_buffer = color;
	}
};

int main(int ac, char *av[])
{
	SDL_Init(SDL_INIT_VIDEO);
	SDL_Surface *screen = SDL_SetVideoMode(320, 240, 32, SDL_SWSURFACE);

	Rasterizer r;
	SDL_SurfaceRenderTarget color_target(screen);
	TestFragmentShader fragment_shader;

	Rasterizer::RenderTarget* rendertargets[] = { &color_target};
	r.rendertargets(1, rendertargets);
	r.fragment_shader(&fragment_shader);
	r.clip_rect(45, 70, 100, 100);

	Rasterizer::Vertex v[3];

	v[0].x = (int)(120.0f * 16.0f);
	v[0].y = (int)(50.0f * 16.0f);
	v[0].z = 0;
	v[0].w = 1 << 16;
	v[0].varyings[0] = 255;
	v[0].varyings[1] = 0;
	v[0].varyings[2] = 0;

	v[1].x = (int)(20.0f  * 16.0f);
	v[1].y = (int)(100.0f * 16.0f);
	v[1].z = 0x7fffffff;
	v[1].w = 1 << 16;
	v[1].varyings[0] = 0;
	v[1].varyings[1] = 255;
	v[1].varyings[2] = 0;

	v[2].x = (int)(150.0f * 16.0f);
	v[2].y = (int)(220.0f * 16.0f);
	v[2].z = 0x7fffffff >> 1;
	v[2].w = 1 << 16;
	v[2].varyings[0] = 0;
	v[2].varyings[1] = 0;
	v[2].varyings[2] = 255;

	SDL_Rect rect;
	rect.x = 45;
	rect.y = 70;
	rect.w = 100;
	rect.h = 100;
	SDL_FillRect(screen, &rect, 0xffffffff);

	r.draw_triangle(v[0], v[1], v[2]);

	SDL_Flip(screen);

	SDL_Event e;
	while (SDL_WaitEvent(&e) && e.type != SDL_QUIT);

	SDL_Quit();
	return 0;
}




I don't include the stdint.h file. You can get it yourself if needed. OK, i now have a domain www.trenki.net where I uploaded a vastly improved version of the renderer. The above code is obsolete, the features are retained. [Edited by - Trenki on August 22, 2007 5:11:28 AM]

Share this post


Link to post
Share on other sites
Quote:
Original post by Trenki
This is done perspectively correct for the corners of each 8x8 block and affine within each block to avoid the costly per pixel divide.

Doesn't this cause errors near the edges of the triangles?

Share this post


Link to post
Share on other sites
Quote:
Original post by C0D1F1ED
Quote:
Original post by Trenki
This is done perspectively correct for the corners of each 8x8 block and affine within each block to avoid the costly per pixel divide.

Doesn't this cause errors near the edges of the triangles?


In the test program I interpolate three colors over the triangle. In the fragment shader it could happen that one of the interpolated rgb components became nagative at the edges of the triangle so I had to clamp them to zero.
But I also looked at the mesa source code and they have the same problem at the edges and AFAIK they are doing the perspective correction per pixel. They have extra code to clamp the colors at the edges of the triangle.

Share this post


Link to post
Share on other sites
Quote:
Original post by Trenki
In the test program I interpolate three colors over the triangle. In the fragment shader it could happen that one of the interpolated rgb components became nagative at the edges of the triangle so I had to clamp them to zero.

Clamping helps for colors, but texture coordinates are unbounded. So at the edges of triangle you could get unwanted wrapping.
Quote:
But I also looked at the mesa source code and they have the same problem at the edges and AFAIK they are doing the perspective correction per pixel. They have extra code to clamp the colors at the edges of the triangle.

Interesting. Is that just to prevent the components from becoming negative due to rounding error then?

Share this post


Link to post
Share on other sites
Quote:
Original post by C0D1F1ED
Quote:
Original post by Trenki
In the test program I interpolate three colors over the triangle. In the fragment shader it could happen that one of the interpolated rgb components became nagative at the edges of the triangle so I had to clamp them to zero.

Clamping helps for colors, but texture coordinates are unbounded. So at the edges of triangle you could get unwanted wrapping.


Yes, this might happen. In December I tried to code a triangle rasterizer following Chris Heckers series on perspective texture mapping. There the varyings are interpolated along the edges but I still got the wrapping effect with the texture coordinates. Maybe I did something wrong... In the end I stopped working on it because I didn't like the code at all and it also was quite involved and not as general as the one above.

Share this post


Link to post
Share on other sites
Quote:
Original post by C0D1F1ED
Quote:
Original post by Eitsch
for me the devmaster link doesn't work. could you tell us what article you mean?
thanks

Advanced Rasterization by Nicolas Capens, also known as c0d1f1ed. [wink]
Huh, that's you?

I'm coding up a softrast based on what you wrote -- though I'm also doing a bit of C++ voodoo with it. That way I don't suffer from the heavy penalty of doing multiple virtual calls for every single fragment (which is what the OP is doing), along with a few other niceties.

Share this post


Link to post
Share on other sites
Quote:
That way I don't suffer from the heavy penalty of doing multiple virtual calls for every single fragment (which is what the OP is doing), along with a few other niceties.


There should be only one virtual function call per fragment and I don't believe this virtual function call is a heavy penalty. I could be using a function pointer instead but do you really believe that would be faster?


Quote:
along with a few other niceties.

Could you please elaborate on this and point out what exacly you mean.

[Edited by - Trenki on May 27, 2007 1:06:21 PM]

Share this post


Link to post
Share on other sites
Have you looked at this paper:
Triangle Scan Conversion using 2D Homogenous Coordinates

It dodges quite nicely many problems related to projection of 3D triangles to 2D, such as, w component being 0. Basicly, you don't need frustum clipping if you use homogenous rasterizer. I definitely recommend you to at least read the paper.

Also one idea worth investigating is using hierarchical z-buffer, which may improve performance especially in case of high depth complexity scenes.

Share this post


Link to post
Share on other sites
Quote:
Original post by Winograd
Have you looked at this paper:
Triangle Scan Conversion using 2D Homogenous Coordinates

It dodges quite nicely many problems related to projection of 3D triangles to 2D, such as, w component being 0. Basicly, you don't need frustum clipping if you use homogenous rasterizer. I definitely recommend you to at least read the paper.

Also one idea worth investigating is using hierarchical z-buffer, which may improve performance especially in case of high depth complexity scenes.


Hi Winograd!

I know of the paper you mentioned and I took a quit look at it but didn't read it all. I didn't understand everything and I also din't get how to actually draw the triangle to the 2d screen at the end + perspective divide for each pixel is too expensive.

Regarding the hierachical z-buffer: I already thought about this and provided a way for the shader to do it. Basically the shader can remember the minimum z of each 8x8 block and than let the rasterizer reject the whole block if it determins that the minimum expected z value for this block is larger.

Share this post


Link to post
Share on other sites
Quote:
Original post by Trenki
There should be only one virtual function call per fragment and I don't believe this virtual function call is a heavy penalty. I could be using a function pointer instead but do you really believe that would be faster?
I do, yes. It's small, but when you're handling things per fragment, that's a lot of calls. (About 29 mil at 800 x 600 at 60fps, and that's with zero overdraw.) Plus I have a few low level ideas I want to play with...
Quote:
Quote:
along with a few other niceties.

Could you please elaborate on this and point out what exacly you mean.
Sorry, it's strictly experimental R&D right now and I would prefer not to say too much until I know for sure what works and what doesn't. Here's a teaser though.
VERTEX_SHADER( Shader )
{
VS_INPUTS (
((Float4, Position, Position))
((Float4, Diffuse, Color))
);

VS_OUTPUTS (
((Float4, Position, Position))
((Float4, Diffuse, Color))
);

VS_Output SimpleVS( const VS_Input& in )
{
VS_Output out;

out.Position = in.Position;
out.Diffuse = in.Diffuse;

return out;
}
} END_VS( Shader );

That's pure C++ code. It should generate vectorized SSE when I'm done.

[Edited by - Promit on May 27, 2007 4:49:07 PM]

Share this post


Link to post
Share on other sites
Quote:
Original post by Trenki
There should be only one virtual function call per fragment and I don't believe this virtual function call is a heavy penalty. I could be using a function pointer instead but do you really believe that would be faster?

Virtual functions are just an abstraction of function pointers. So it wouldn't be faster to use a function pointer explicitely. However, argument passing is quite expensive. You might be able to speed things up a little by making the arguments class members and sharing them with the shader routine. Beware of turning things into spaghettic code though...

Share this post


Link to post
Share on other sites
Quote:
Original post by Winograd
Basicly, you don't need frustum clipping if you use homogenous rasterizer.

That's a nice property for a hardware implementation, but in software clipping is quite simple and fast. The homogenous rasterizer needs extra work per pixel, which makes it less attractive for software.
Quote:
Also one idea worth investigating is using hierarchical z-buffer, which may improve performance especially in case of high depth complexity scenes.

Yeah, my implementation with 8x8 blocks can be used directly with a hierarchical z-buffer. In my experience it's not worth it when working with low resolutions though (typically the case for a software renderer), but your mileage may vary.

Share this post


Link to post
Share on other sites
Quote:
Original post by Promit
Sorry, it's strictly experimental R&D right now and I would prefer not to say too much until I know for sure what works and what doesn't. Here's a teaser though.*** Source Snippet Removed ***That's pure C++ code. It should generate vectorized SSE when I'm done.

Have you looked at Sh yet? It's capable of writing out C code that can be compiled at run-time with GCC or ICC.

What back-end are you using to generate SSE code?

Share this post


Link to post
Share on other sites
Question to C0D1F1ED: As you pointed out I could get an undesired wrapping effect for the texture coordinates at the edges of the triangles. I have given it some thought but could not come up with a satisfactory solution. What would you suggest without requiring the expensive perspective correction and retaining the overall design?

Share this post


Link to post
Share on other sites
Quote:
Original post by Trenki
As you pointed out I could get an undesired wrapping effect for the texture coordinates at the edges of the triangles. I have given it some thought but could not come up with a satisfactory solution. What would you suggest without requiring the expensive perspective correction and retaining the overall design?

For fully covered tiles, keep using linear interpolation. For partially covered tiles, use per-pixel perspective correction.

To avoid even more perspective correction, you can detect which triangles are either small enough or 'flat' enough not to require perspective correction at all...

But beware of the law of premature optimization! This is going to complicate your design, and could make it quite complicated to maintain the code. So only do it if you really need it and everything else is functional.

The wrapping is not that terrible, so depending on the needs of your projects you might not need to solve it at all.

Share this post


Link to post
Share on other sites
Quote:
Original post by C0D1F1ED
Quote:
Original post by Winograd
Basicly, you don't need frustum clipping if you use homogenous rasterizer.

That's a nice property for a hardware implementation, but in software clipping is quite simple and fast. The homogenous rasterizer needs extra work per pixel, which makes it less attractive for software.


Clipping is quite fast on hardware also ;) Well I guess your point is that on hardware there is no extra per pixel cost except in the number of gates. I haven't implemented such rasterizer but at quick glance it seems one could apply "DDA"-like algorithm which would account to basicly 4 additions per pixel of which one is used for perspective division.

Share this post


Link to post
Share on other sites
I've now coded up a vertex transformation and clipping pipeline and tested the whole stuff with a cube. It runs fine for five seconds and then I get an integer division by 0 in the rasterizer.

I could trace it to this line
int w10 = fixdiv<16>(1 << 16, solve_plane(xx1, y0, wPlane));

Apparently solve_plane returns 0 here. I checked with my calculator using the values VS studio gave me but if I did the calculations correctly it should really ba returning 1. Still this is a problem since such things should not happen.

The problems with the triangles at the top of the cube when it is viewed at a shallow angle (y coordinates of the corners: 243 244 246).

Does anyone have suggestions on how to counter this problem?

Share this post


Link to post
Share on other sites
Quote:
Original post by Trenki
I checked with my calculator using the values VS studio gave me but if I did the calculations correctly it should really ba returning 1.

Did you enable break on exceptions in the Debug > Exceptions... menu? When the exception occurs, break, and then place the 'yellow arrow' at the start of the calculation (you might need to first step out of the function, then move the arrow). This way you can interactively follow the calculations. Pressing Alt+F8 will show the disassembly so you can follow one instruction at a time. Or you can split your C++ calculation up into elemental computations so you can follow line by line.

That should quickly reveal the cause of the error...

Share this post


Link to post
Share on other sites
The division by zero exception occurred and visual studio pointed me right at the location where that happend. The value solve_plane returned really was zero, so that resulted in the following division by 0.

The thing is that I compute 1/w for the corners of each 8x8 block using the plane equation calculated earlier. From this i compute W by taking the reciprocal. Since the corners of the block can lie outside the triangle I also can get values that I would not get if I stayed inside the triangle all the time.

I fixed it by simply setting the value to 1 if it was 0. Now the demo program which animates a simple cube runs.

Do you know where the profiler in Visual Studio is? Currently I use Orcas Beta 1.

Share this post


Link to post
Share on other sites
Hi!

I have finished my vertex and clipping pipeline, so I could do some speed tests with my rasterizer. I get 600fps at 320x240 and ~60fps at 1024x768 on my AMD Athlon64 3500+ (2.2 Ghz) for a scene with a single gouraud shaded cube. The cube constanly fills ~1/9 of the screen area. As I don't have any numbers to compare this to I don't know how good/bad this is.

Thinking back to the days of DOS where I played Fatal Racing at 640x480 on my Pentium 166 which ran at ~20-30fps and filled the whole screen and also had texture mapping my rasterizer seems slow. What do you guys think?

I also did some profiling with gprof:

% cumulative self
time seconds seconds calls name
72.20 4.26 4.26 31899 Rasterizer::draw_triangle
14.07 5.09 0.83 __divdi3
12.54 5.83 0.74 47792689 TestFragmentShader::shade

Even though shade is called 47 million times it requires less time than draw_triangle. Probably because the interpolation of the varyings happens in draw_triangle. I don't know how much the virtual function call per pixel affects the performance and wether the time spent on the call is added to the draw_triangle total time or to the shade time. I will try to remove the parameters from the shade function and find anonther way to let the shader know about the required values. Maybe that speeds things up a little.

[Edited by - Trenki on May 29, 2007 12:56:49 PM]

Share this post


Link to post
Share on other sites
Quote:
Original post by Trenki
I have finished my vertex and clipping pipeline, so I could do some speed tests with my rasterizer.

I recommend you to run some simple sweep tests. For example, keep the total rasterized pixels constant by rectangular area with strip of triangles. Vary the number of triangles per frame linearly, starting from 1 or 2 and going up. Notice that you will first be fill limited, and thus increasing triangle count does not have much effect. When the graph starts to climb linearly you are geometry limited and the slope gives you your peak triangle rate.

In second test, keep the triangle count constant, but vary the number of rasterized pixels. The slope will give you the peak pixel rate.

You can calculate the approximate number clks used per pixel basis and overhead caused by geometry processing. Ideally geometry would be processed in parallel with rasterization.

Share this post


Link to post
Share on other sites
Quote:
Original post by Trenki
I get 600fps at 320x240 and ~60fps at 1024x768 on my AMD Athlon64 3500+ (2.2 Ghz) for a scene with a single gouraud shaded cube. The cube constanly fills ~1/9 of the screen area. As I don't have any numbers to compare this to I don't know how good/bad this is.

What you really should be questioning is what are your own goals? Increasing performance at this point is without a doubt going to make your software hard to maintain. So if you want to add texture mapping and things like that, add it first. As long as things are interactive enough to test, performance is really fine. Once all functionality you need is implemented you can concentrate on the real bottlenecks. It's very likely that texturing will be a major new bottneck, so much of the work you'd currenly do on the gouraud shading would be largely pointless.

Also, if you really target the GP2X then you should only look at how it performs on that. At the resolution of 320x240 it might not even be worth it to use 8x8 pixel tiles. Different architectures have different needs. Maybe it's bandwidth limited and you should really concentrate on addressing that first...

So the best advice I can give you is to stop programming and start developing. Write down your goals so you have something to concentrate on. That way we also know what advice to give you, instead of sending you in the wrong directions. Trust me, pinpointing your goals is an extremely crucial step towards a succesful project.

Share this post


Link to post
Share on other sites

This topic is 3843 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Sign in to follow this  

  • Forum Statistics

    • Total Topics
      628636
    • Total Posts
      2983966
  • Similar Content

    • By HD86
      I have a vertex buffer on a default heap. I need a CPU pointer to that buffer in order to loop through the vertices and change one value in some vertices (the color value). In the past this was possible by creating the buffer with the flag D3DUSAGE_DYNAMIC/D3D11_USAGE_DYNAMIC and using IDirect3DVertexBuffer9::Lock or ID3D11DeviceContext::Map to get a pointer.
      What is the correct way to do the same in DX 12? As far as I understand, the method ID3D12Resource::Map cannot be used on a default heap because default heaps cannot be accessed directly from the CPU. The documentation says that upload heaps are intended for CPU-write-once, GPU-read-once usage, so I don't think these are equivalent to the "dynamic" buffers. Is the readback heap equivalent to what was called a dynamic buffer? Or should I create a custom heap?
      I am thinking to do the following:
      -Create a temporary readback heap.
      -Copy the data from the default heap to the readback heap using UpdateSubresources.
      -Get a CPU pointer to the readback heap using Map and edit the data.
      -Copy the data back to the default heap using UpdateSubresources.
      What do you think about this?
    • By lubbe75
      Being new to DirectX 12 I am looking for examples on how to use threading. I have done lots of OpenGL in the past and some DirectX, but with DX12 the threading magic is gone and I understand that threading is crucial to get good performance. In my project I currently have one thread doing it all. I have one command list, one command allocator, one bundle and one bundle allocator. I also have a million triangles, so it's about time that I start doing this.
      How do I split things up? How many threads should I use? How many command lists and allocators?
      I realize this is a beginner's question , but I have to begin somewhere. I would be grateful if someone could point me in a direction where I could find a simple code sample, tutorial or something similar. Thanks!
    • By turanszkij
      I am porting a DX11 game engine to DX12, and have some questions about root signature usage. The engine clearly defines a global bind slot range for resources, to avoid using shader reflection. This means, that both the application, both the shaders explicitly know the binding slots for everything (constant buffers, textures, etc.). This would trivially translate to a global root signature in DX12, and I have created it as a big root signature like the following: There are 12 descriptor tables, so for every shader stage there are a sampler descriptor table and a CBV_SRV_UAV table. I though about keeping a "staging" descriptor heap, to which I would place descriptors at resource load time. Then when calling PSSetShaderResources(), which is a DX11-style function, the function would fill out a GPU-visible descriptor heap which is created to match the layout of the root signature tables. The nice thing about is that each frame I would only have a single call to set the root signature and to set root signature tables. The command list would just fill it with descriptors before rendering.
      I've found out however, that filling out descriptor tables can only be done on the CPU timeline with ID3D12Device::CopyDescriptors(). This complicates things and I can only think of populating the GPU descriptor heap, and after each draw call, I would need to copy the whole table and CopyDescriptors() again when there is a new call to PSSetShaderResources(). Am I on the right path if I also want to keep DX11-style bindings? 
      I also thought about each shader having a custom root signature and only ever copying the descriptors which are required by the shader, but in this case I will need to rebind the whole root signature each time a shader changes. I would also have to do a shader reflection step and modify my shader-asset pipeline (because I don't want to depend on d3dcompiler when I run the engine) which I really want to avoid.
      Which would be the preferred way? I read the Nvidia Do's and Don'ts but they say these which are contradicting in my case:
      Minimize the number of Root Signature changes Don’t bloat your root signature and descriptor tables to be able to reuse them UPDATE: The second way (unique root signature with PSO) I would also have to bind resources strictly after I have set the right PSO, because the table is now unique. 
    • By DiligentDev
      This article uses material originally posted on Diligent Graphics web site.
      Introduction
      Graphics APIs have come a long way from small set of basic commands allowing limited control of configurable stages of early 3D accelerators to very low-level programming interfaces exposing almost every aspect of the underlying graphics hardware. Next-generation APIs, Direct3D12 by Microsoft and Vulkan by Khronos are relatively new and have only started getting widespread adoption and support from hardware vendors, while Direct3D11 and OpenGL are still considered industry standard. New APIs can provide substantial performance and functional improvements, but may not be supported by older hardware. An application targeting wide range of platforms needs to support Direct3D11 and OpenGL. New APIs will not give any advantage when used with old paradigms. It is totally possible to add Direct3D12 support to an existing renderer by implementing Direct3D11 interface through Direct3D12, but this will give zero benefits. Instead, new approaches and rendering architectures that leverage flexibility provided by the next-generation APIs are expected to be developed.
      There are at least four APIs (Direct3D11, Direct3D12, OpenGL/GLES, Vulkan, plus Apple's Metal for iOS and osX platforms) that a cross-platform 3D application may need to support. Writing separate code paths for all APIs is clearly not an option for any real-world application and the need for a cross-platform graphics abstraction layer is evident. The following is the list of requirements that I believe such layer needs to satisfy:
      Lightweight abstractions: the API should be as close to the underlying native APIs as possible to allow an application leverage all available low-level functionality. In many cases this requirement is difficult to achieve because specific features exposed by different APIs may vary considerably. Low performance overhead: the abstraction layer needs to be efficient from performance point of view. If it introduces considerable amount of overhead, there is no point in using it. Convenience: the API needs to be convenient to use. It needs to assist developers in achieving their goals not limiting their control of the graphics hardware. Multithreading: ability to efficiently parallelize work is in the core of Direct3D12 and Vulkan and one of the main selling points of the new APIs. Support for multithreading in a cross-platform layer is a must. Extensibility: no matter how well the API is designed, it still introduces some level of abstraction. In some cases the most efficient way to implement certain functionality is to directly use native API. The abstraction layer needs to provide seamless interoperability with the underlying native APIs to provide a way for the app to add features that may be missing. Diligent Engine is designed to solve these problems. Its main goal is to take advantages of the next-generation APIs such as Direct3D12 and Vulkan, but at the same time provide support for older platforms via Direct3D11, OpenGL and OpenGLES. Diligent Engine exposes common C++ front-end for all supported platforms and provides interoperability with underlying native APIs. It also supports integration with Unity and is designed to be used as graphics subsystem in a standalone game engine, Unity native plugin or any other 3D application. Full source code is available for download at GitHub and is free to use.
      Overview
      Diligent Engine API takes some features from Direct3D11 and Direct3D12 as well as introduces new concepts to hide certain platform-specific details and make the system easy to use. It contains the following main components:
      Render device (IRenderDevice  interface) is responsible for creating all other objects (textures, buffers, shaders, pipeline states, etc.).
      Device context (IDeviceContext interface) is the main interface for recording rendering commands. Similar to Direct3D11, there are immediate context and deferred contexts (which in Direct3D11 implementation map directly to the corresponding context types). Immediate context combines command queue and command list recording functionality. It records commands and submits the command list for execution when it contains sufficient number of commands. Deferred contexts are designed to only record command lists that can be submitted for execution through the immediate context.
      An alternative way to design the API would be to expose command queue and command lists directly. This approach however does not map well to Direct3D11 and OpenGL. Besides, some functionality (such as dynamic descriptor allocation) can be much more efficiently implemented when it is known that a command list is recorded by a certain deferred context from some thread.
      The approach taken in the engine does not limit scalability as the application is expected to create one deferred context per thread, and internally every deferred context records a command list in lock-free fashion. At the same time this approach maps well to older APIs.
      In current implementation, only one immediate context that uses default graphics command queue is created. To support multiple GPUs or multiple command queue types (compute, copy, etc.), it is natural to have one immediate contexts per queue. Cross-context synchronization utilities will be necessary.
      Swap Chain (ISwapChain interface). Swap chain interface represents a chain of back buffers and is responsible for showing the final rendered image on the screen.
      Render device, device contexts and swap chain are created during the engine initialization.
      Resources (ITexture and IBuffer interfaces). There are two types of resources - textures and buffers. There are many different texture types (2D textures, 3D textures, texture array, cubmepas, etc.) that can all be represented by ITexture interface.
      Resources Views (ITextureView and IBufferView interfaces). While textures and buffers are mere data containers, texture views and buffer views describe how the data should be interpreted. For instance, a 2D texture can be used as a render target for rendering commands or as a shader resource.
      Pipeline State (IPipelineState interface). GPU pipeline contains many configurable stages (depth-stencil, rasterizer and blend states, different shader stage, etc.). Direct3D11 uses coarse-grain objects to set all stage parameters at once (for instance, a rasterizer object encompasses all rasterizer attributes), while OpenGL contains myriad functions to fine-grain control every individual attribute of every stage. Both methods do not map very well to modern graphics hardware that combines all states into one monolithic state under the hood. Direct3D12 directly exposes pipeline state object in the API, and Diligent Engine uses the same approach.
      Shader Resource Binding (IShaderResourceBinding interface). Shaders are programs that run on the GPU. Shaders may access various resources (textures and buffers), and setting correspondence between shader variables and actual resources is called resource binding. Resource binding implementation varies considerably between different API. Diligent Engine introduces a new object called shader resource binding that encompasses all resources needed by all shaders in a certain pipeline state.
      API Basics
      Creating Resources
      Device resources are created by the render device. The two main resource types are buffers, which represent linear memory, and textures, which use memory layouts optimized for fast filtering. Graphics APIs usually have a native object that represents linear buffer. Diligent Engine uses IBuffer interface as an abstraction for a native buffer. To create a buffer, one needs to populate BufferDesc structure and call IRenderDevice::CreateBuffer() method as in the following example:
      BufferDesc BuffDesc; BufferDesc.Name = "Uniform buffer"; BuffDesc.BindFlags = BIND_UNIFORM_BUFFER; BuffDesc.Usage = USAGE_DYNAMIC; BuffDesc.uiSizeInBytes = sizeof(ShaderConstants); BuffDesc.CPUAccessFlags = CPU_ACCESS_WRITE; m_pDevice->CreateBuffer( BuffDesc, BufferData(), &m_pConstantBuffer ); While there is usually just one buffer object, different APIs use very different approaches to represent textures. For instance, in Direct3D11, there are ID3D11Texture1D, ID3D11Texture2D, and ID3D11Texture3D objects. In OpenGL, there is individual object for every texture dimension (1D, 2D, 3D, Cube), which may be a texture array, which may also be multisampled (i.e. GL_TEXTURE_2D_MULTISAMPLE_ARRAY). As a result there are nine different GL texture types that Diligent Engine may create under the hood. In Direct3D12, there is only one resource interface. Diligent Engine hides all these details in ITexture interface. There is only one  IRenderDevice::CreateTexture() method that is capable of creating all texture types. Dimension, format, array size and all other parameters are specified by the members of the TextureDesc structure:
      TextureDesc TexDesc; TexDesc.Name = "My texture 2D"; TexDesc.Type = TEXTURE_TYPE_2D; TexDesc.Width = 1024; TexDesc.Height = 1024; TexDesc.Format = TEX_FORMAT_RGBA8_UNORM; TexDesc.Usage = USAGE_DEFAULT; TexDesc.BindFlags = BIND_SHADER_RESOURCE | BIND_RENDER_TARGET | BIND_UNORDERED_ACCESS; TexDesc.Name = "Sample 2D Texture"; m_pRenderDevice->CreateTexture( TexDesc, TextureData(), &m_pTestTex ); If native API supports multithreaded resource creation, textures and buffers can be created by multiple threads simultaneously.
      Interoperability with native API provides access to the native buffer/texture objects and also allows creating Diligent Engine objects from native handles. It allows applications seamlessly integrate native API-specific code with Diligent Engine.
      Next-generation APIs allow fine level-control over how resources are allocated. Diligent Engine does not currently expose this functionality, but it can be added by implementing IResourceAllocator interface that encapsulates specifics of resource allocation and providing this interface to CreateBuffer() or CreateTexture() methods. If null is provided, default allocator should be used.
      Initializing the Pipeline State
      As it was mentioned earlier, Diligent Engine follows next-gen APIs to configure the graphics/compute pipeline. One big Pipelines State Object (PSO) encompasses all required states (all shader stages, input layout description, depth stencil, rasterizer and blend state descriptions etc.). This approach maps directly to Direct3D12/Vulkan, but is also beneficial for older APIs as it eliminates pipeline misconfiguration errors. With many individual calls tweaking various GPU pipeline settings it is very easy to forget to set one of the states or assume the stage is already properly configured when in fact it is not. Using pipeline state object helps avoid these problems as all stages are configured at once.
      Creating Shaders
      While in earlier APIs shaders were bound separately, in the next-generation APIs as well as in Diligent Engine shaders are part of the pipeline state object. The biggest challenge when authoring shaders is that Direct3D and OpenGL/Vulkan use different shader languages (while Apple uses yet another language in their Metal API). Maintaining two versions of every shader is not an option for real applications and Diligent Engine implements shader source code converter that allows shaders authored in HLSL to be translated to GLSL. To create a shader, one needs to populate ShaderCreationAttribs structure. SourceLanguage member of this structure tells the system which language the shader is authored in:
      SHADER_SOURCE_LANGUAGE_DEFAULT - The shader source language matches the underlying graphics API: HLSL for Direct3D11/Direct3D12 mode, and GLSL for OpenGL and OpenGLES modes. SHADER_SOURCE_LANGUAGE_HLSL - The shader source is in HLSL. For OpenGL and OpenGLES modes, the source code will be converted to GLSL. SHADER_SOURCE_LANGUAGE_GLSL - The shader source is in GLSL. There is currently no GLSL to HLSL converter, so this value should only be used for OpenGL and OpenGLES modes. There are two ways to provide the shader source code. The first way is to use Source member. The second way is to provide a file path in FilePath member. Since the engine is entirely decoupled from the platform and the host file system is platform-dependent, the structure exposes pShaderSourceStreamFactory member that is intended to provide the engine access to the file system. If FilePath is provided, shader source factory must also be provided. If the shader source contains any #include directives, the source stream factory will also be used to load these files. The engine provides default implementation for every supported platform that should be sufficient in most cases. Custom implementation can be provided when needed.
      When sampling a texture in a shader, the texture sampler was traditionally specified as separate object that was bound to the pipeline at run time or set as part of the texture object itself. However, in most cases it is known beforehand what kind of sampler will be used in the shader. Next-generation APIs expose new type of sampler called static sampler that can be initialized directly in the pipeline state. Diligent Engine exposes this functionality: when creating a shader, textures can be assigned static samplers. If static sampler is assigned, it will always be used instead of the one initialized in the texture shader resource view. To initialize static samplers, prepare an array of StaticSamplerDesc structures and initialize StaticSamplers and NumStaticSamplers members. Static samplers are more efficient and it is highly recommended to use them whenever possible. On older APIs, static samplers are emulated via generic sampler objects.
      The following is an example of shader initialization:
      ShaderCreationAttribs Attrs; Attrs.Desc.Name = "MyPixelShader"; Attrs.FilePath = "MyShaderFile.fx"; Attrs.SearchDirectories = "shaders;shaders\\inc;"; Attrs.EntryPoint = "MyPixelShader"; Attrs.Desc.ShaderType = SHADER_TYPE_PIXEL; Attrs.SourceLanguage = SHADER_SOURCE_LANGUAGE_HLSL; BasicShaderSourceStreamFactory BasicSSSFactory(Attrs.SearchDirectories); Attrs.pShaderSourceStreamFactory = &BasicSSSFactory; ShaderVariableDesc ShaderVars[] = {     {"g_StaticTexture", SHADER_VARIABLE_TYPE_STATIC},     {"g_MutableTexture", SHADER_VARIABLE_TYPE_MUTABLE},     {"g_DynamicTexture", SHADER_VARIABLE_TYPE_DYNAMIC} }; Attrs.Desc.VariableDesc = ShaderVars; Attrs.Desc.NumVariables = _countof(ShaderVars); Attrs.Desc.DefaultVariableType = SHADER_VARIABLE_TYPE_STATIC; StaticSamplerDesc StaticSampler; StaticSampler.Desc.MinFilter = FILTER_TYPE_LINEAR; StaticSampler.Desc.MagFilter = FILTER_TYPE_LINEAR; StaticSampler.Desc.MipFilter = FILTER_TYPE_LINEAR; StaticSampler.TextureName = "g_MutableTexture"; Attrs.Desc.NumStaticSamplers = 1; Attrs.Desc.StaticSamplers = &StaticSampler; ShaderMacroHelper Macros; Macros.AddShaderMacro("USE_SHADOWS", 1); Macros.AddShaderMacro("NUM_SHADOW_SAMPLES", 4); Macros.Finalize(); Attrs.Macros = Macros; RefCntAutoPtr<IShader> pShader; m_pDevice->CreateShader( Attrs, &pShader );
      Creating the Pipeline State Object
      After all required shaders are created, the rest of the fields of the PipelineStateDesc structure provide depth-stencil, rasterizer, and blend state descriptions, the number and format of render targets, input layout format, etc. For instance, rasterizer state can be described as follows:
      PipelineStateDesc PSODesc; RasterizerStateDesc &RasterizerDesc = PSODesc.GraphicsPipeline.RasterizerDesc; RasterizerDesc.FillMode = FILL_MODE_SOLID; RasterizerDesc.CullMode = CULL_MODE_NONE; RasterizerDesc.FrontCounterClockwise = True; RasterizerDesc.ScissorEnable = True; RasterizerDesc.AntialiasedLineEnable = False; Depth-stencil and blend states are defined in a similar fashion.
      Another important thing that pipeline state object encompasses is the input layout description that defines how inputs to the vertex shader, which is the very first shader stage, should be read from the memory. Input layout may define several vertex streams that contain values of different formats and sizes:
      // Define input layout InputLayoutDesc &Layout = PSODesc.GraphicsPipeline.InputLayout; LayoutElement TextLayoutElems[] = {     LayoutElement( 0, 0, 3, VT_FLOAT32, False ),     LayoutElement( 1, 0, 4, VT_UINT8, True ),     LayoutElement( 2, 0, 2, VT_FLOAT32, False ), }; Layout.LayoutElements = TextLayoutElems; Layout.NumElements = _countof( TextLayoutElems ); Finally, pipeline state defines primitive topology type. When all required members are initialized, a pipeline state object can be created by IRenderDevice::CreatePipelineState() method:
      // Define shader and primitive topology PSODesc.GraphicsPipeline.PrimitiveTopologyType = PRIMITIVE_TOPOLOGY_TYPE_TRIANGLE; PSODesc.GraphicsPipeline.pVS = pVertexShader; PSODesc.GraphicsPipeline.pPS = pPixelShader; PSODesc.Name = "My pipeline state"; m_pDev->CreatePipelineState(PSODesc, &m_pPSO); When PSO object is bound to the pipeline, the engine invokes all API-specific commands to set all states specified by the object. In case of Direct3D12 this maps directly to setting the D3D12 PSO object. In case of Direct3D11, this involves setting individual state objects (such as rasterizer and blend states), shaders, input layout etc. In case of OpenGL, this requires a number of fine-grain state tweaking calls. Diligent Engine keeps track of currently bound states and only calls functions to update these states that have actually changed.
      Binding Shader Resources
      Direct3D11 and OpenGL utilize fine-grain resource binding models, where an application binds individual buffers and textures to certain shader or program resource binding slots. Direct3D12 uses a very different approach, where resource descriptors are grouped into tables, and an application can bind all resources in the table at once by setting the table in the command list. Resource binding model in Diligent Engine is designed to leverage this new method. It introduces a new object called shader resource binding that encapsulates all resource bindings required for all shaders in a certain pipeline state. It also introduces the classification of shader variables based on the frequency of expected change that helps the engine group them into tables under the hood:
      Static variables (SHADER_VARIABLE_TYPE_STATIC) are variables that are expected to be set only once. They may not be changed once a resource is bound to the variable. Such variables are intended to hold global constants such as camera attributes or global light attributes constant buffers. Mutable variables (SHADER_VARIABLE_TYPE_MUTABLE) define resources that are expected to change on a per-material frequency. Examples may include diffuse textures, normal maps etc. Dynamic variables (SHADER_VARIABLE_TYPE_DYNAMIC) are expected to change frequently and randomly. Shader variable type must be specified during shader creation by populating an array of ShaderVariableDesc structures and initializing ShaderCreationAttribs::Desc::VariableDesc and ShaderCreationAttribs::Desc::NumVariables members (see example of shader creation above).
      Static variables cannot be changed once a resource is bound to the variable. They are bound directly to the shader object. For instance, a shadow map texture is not expected to change after it is created, so it can be bound directly to the shader:
      PixelShader->GetShaderVariable( "g_tex2DShadowMap" )->Set( pShadowMapSRV ); Mutable and dynamic variables are bound via a new Shader Resource Binding object (SRB) that is created by the pipeline state (IPipelineState::CreateShaderResourceBinding()):
      m_pPSO->CreateShaderResourceBinding(&m_pSRB); Note that an SRB is only compatible with the pipeline state it was created from. SRB object inherits all static bindings from shaders in the pipeline, but is not allowed to change them.
      Mutable resources can only be set once for every instance of a shader resource binding. Such resources are intended to define specific material properties. For instance, a diffuse texture for a specific material is not expected to change once the material is defined and can be set right after the SRB object has been created:
      m_pSRB->GetVariable(SHADER_TYPE_PIXEL, "tex2DDiffuse")->Set(pDiffuseTexSRV); In some cases it is necessary to bind a new resource to a variable every time a draw command is invoked. Such variables should be labeled as dynamic, which will allow setting them multiple times through the same SRB object:
      m_pSRB->GetVariable(SHADER_TYPE_VERTEX, "cbRandomAttribs")->Set(pRandomAttrsCB); Under the hood, the engine pre-allocates descriptor tables for static and mutable resources when an SRB objcet is created. Space for dynamic resources is dynamically allocated at run time. Static and mutable resources are thus more efficient and should be used whenever possible.
      As you can see, Diligent Engine does not expose low-level details of how resources are bound to shader variables. One reason for this is that these details are very different for various APIs. The other reason is that using low-level binding methods is extremely error-prone: it is very easy to forget to bind some resource, or bind incorrect resource such as bind a buffer to the variable that is in fact a texture, especially during shader development when everything changes fast. Diligent Engine instead relies on shader reflection system to automatically query the list of all shader variables. Grouping variables based on three types mentioned above allows the engine to create optimized layout and take heavy lifting of matching resources to API-specific resource location, register or descriptor in the table.
      This post gives more details about the resource binding model in Diligent Engine.
      Setting the Pipeline State and Committing Shader Resources
      Before any draw or compute command can be invoked, the pipeline state needs to be bound to the context:
      m_pContext->SetPipelineState(m_pPSO); Under the hood, the engine sets the internal PSO object in the command list or calls all the required native API functions to properly configure all pipeline stages.
      The next step is to bind all required shader resources to the GPU pipeline, which is accomplished by IDeviceContext::CommitShaderResources() method:
      m_pContext->CommitShaderResources(m_pSRB, COMMIT_SHADER_RESOURCES_FLAG_TRANSITION_RESOURCES); The method takes a pointer to the shader resource binding object and makes all resources the object holds available for the shaders. In the case of D3D12, this only requires setting appropriate descriptor tables in the command list. For older APIs, this typically requires setting all resources individually.
      Next-generation APIs require the application to track the state of every resource and explicitly inform the system about all state transitions. For instance, if a texture was used as render target before, while the next draw command is going to use it as shader resource, a transition barrier needs to be executed. Diligent Engine does the heavy lifting of state tracking.  When CommitShaderResources() method is called with COMMIT_SHADER_RESOURCES_FLAG_TRANSITION_RESOURCES flag, the engine commits and transitions resources to correct states at the same time. Note that transitioning resources does introduce some overhead. The engine tracks state of every resource and it will not issue the barrier if the state is already correct. But checking resource state is an overhead that can sometimes be avoided. The engine provides IDeviceContext::TransitionShaderResources() method that only transitions resources:
      m_pContext->TransitionShaderResources(m_pPSO, m_pSRB); In some scenarios it is more efficient to transition resources once and then only commit them.
      Invoking Draw Command
      The final step is to set states that are not part of the PSO, such as render targets, vertex and index buffers. Diligent Engine uses Direct3D11-syle API that is translated to other native API calls under the hood:
      ITextureView *pRTVs[] = {m_pRTV}; m_pContext->SetRenderTargets(_countof( pRTVs ), pRTVs, m_pDSV); // Clear render target and depth buffer const float zero[4] = {0, 0, 0, 0}; m_pContext->ClearRenderTarget(nullptr, zero); m_pContext->ClearDepthStencil(nullptr, CLEAR_DEPTH_FLAG, 1.f); // Set vertex and index buffers IBuffer *buffer[] = {m_pVertexBuffer}; Uint32 offsets[] = {0}; Uint32 strides[] = {sizeof(MyVertex)}; m_pContext->SetVertexBuffers(0, 1, buffer, strides, offsets, SET_VERTEX_BUFFERS_FLAG_RESET); m_pContext->SetIndexBuffer(m_pIndexBuffer, 0); Different native APIs use various set of function to execute draw commands depending on command details (if the command is indexed, instanced or both, what offsets in the source buffers are used etc.). For instance, there are 5 draw commands in Direct3D11 and more than 9 commands in OpenGL with something like glDrawElementsInstancedBaseVertexBaseInstance not uncommon. Diligent Engine hides all details with single IDeviceContext::Draw() method that takes takes DrawAttribs structure as an argument. The structure members define all attributes required to perform the command (primitive topology, number of vertices or indices, if draw call is indexed or not, if draw call is instanced or not, if draw call is indirect or not, etc.). For example:
      DrawAttribs attrs; attrs.IsIndexed = true; attrs.IndexType = VT_UINT16; attrs.NumIndices = 36; attrs.Topology = PRIMITIVE_TOPOLOGY_TRIANGLE_LIST; pContext->Draw(attrs); For compute commands, there is IDeviceContext::DispatchCompute() method that takes DispatchComputeAttribs structure that defines compute grid dimension.
      Source Code
      Full engine source code is available on GitHub and is free to use. The repository contains two samples, asteroids performance benchmark and example Unity project that uses Diligent Engine in native plugin.
      AntTweakBar sample is Diligent Engine’s “Hello World” example.

       
      Atmospheric scattering sample is a more advanced example. It demonstrates how Diligent Engine can be used to implement various rendering tasks: loading textures from files, using complex shaders, rendering to multiple render targets, using compute shaders and unordered access views, etc.

      Asteroids performance benchmark is based on this demo developed by Intel. It renders 50,000 unique textured asteroids and allows comparing performance of Direct3D11 and Direct3D12 implementations. Every asteroid is a combination of one of 1000 unique meshes and one of 10 unique textures.

      Finally, there is an example project that shows how Diligent Engine can be integrated with Unity.

      Future Work
      The engine is under active development. It currently supports Windows desktop, Universal Windows and Android platforms. Direct3D11, Direct3D12, OpenGL/GLES backends are now feature complete. Vulkan backend is coming next, and support for more platforms is planned.
    • By turanszkij
      I am doing a DX12 graphics wrapper, and I would like to update constant buffers. I found the ID3D12GraphicsCommandList2::WriteBufferImmediate method, which is apparently available from a Windows 10 Creators update only. I couldn't really find any info about this (and couldn't try it yet), am I correct to assume this would be useful for writing to constant buffers without much need to do synchronization? It seems to me like this method copies data to the command list itself and then that data will be copied into the DEFAULT resource address which I provided? The only synchronization needed here would be transition barriers to COPY_DEST before WriteBufferImmediate() and back to GENERIC_READ afterwards? I could be totally off though, I'm still wrapping my head around a lot of things.
      What other use cases would this method allow for?
  • Popular Now