fractorium/Source/EmberTester/EmberTester.cpp

2563 lines
74 KiB
C++
Raw Normal View History

#include "EmberCommonPch.h"
#include "EmberTester.h"
#include "JpegUtils.h"
#include <queue>
#include <list>
#include <deque>
#ifdef __APPLE__
#include <OpenEXR/ImfRgbaFile.h>
#include <OpenEXR/ImfStringAttribute.h>
#include <OpenEXR/half.h>
#else
#include <ImfRgbaFile.h>
#include <ImfStringAttribute.h>
#include <half.h>
#endif
Features: --Add support for Exr files which use 32-bit floats for each RGBA channel. Seems to come out too washed out. --Allow for clearing an individual color curve. --Allow for saving multiple image types in EmberRender and EmberAnimate. All writes are threaded. --Remove --bpc command line argument. Add format png16 as a replacement. --Remove --enable_jpg_comments and --enable_png_comments command line arguments, and replace them with --enable_comments which applies to jpg, png and exr. --Add menu items to variations and affine spinners which allow for easy entry of specific numeric values like pi. --Make final render dialog be wider rather than so tall. Bug fixes: --Fix some OpenCL compile errors on Mac. --Remove ability to save bitmap files on all platforms but Windows. Code changes: --New dependency on OpenEXR. --Allow Curves class to interact with objects of a different template type. --Make m_Curves member of Ember always use float as template type. --Set the length of the curves array to always be 2^17 which should offer enough precision with new 32-bit float pixel types. --Set pixel types to always be 32-bit float. This results in a major reduction of code in the final accumulation part of Renderer.h/cpp. --Remove corresponding code from RendererCL and FinalAccumOpenCLKernelCreator. --Remove Transparency, NumChannels and BytesPerPixel setters from Renderer.h/cpp. --Add new global functions to format final image buffers and place all alpha calculation and scaling code in them. --Blending is no longer needed in OpenGLWidget because of the new pixel type. --Make new class, AffineDoubleSpinBox. --Attempt to make file save dialog code work the same on all OSes. --Remove some unused functions.
2017-07-22 16:43:35 -04:00
//#include <ImfStringAttribute.h>
//#include <ImfMatrixAttribute.h>
//#include <ImfArray.h>
//#include <ImfChannelList.h>
//#include "drawImage.h"
//
//#include <iostream>
//#include <algorithm>
//
//
//#include <ImfNamespace.h>
//
//namespace IMF = Imf;
//
using namespace Imf;
using namespace Imath;
/// <summary>
/// EmberTester is a scratch area used for on the fly testing.
/// It may become a more formalized automated testing system
/// in the future. At the moment it isn't expected to build or
/// give any useful insight into the workings of Ember.
/// </summary>
using namespace EmberNs;
--User changes -Add animation sequence creation to Fractorium. -Add two new options to EmberGenome which are used when generating an animation sequence.: --startcount: Add this number to the filename of each flame. --padding: Override the automatically calculated amount of padding zeroes added to each filename. --Bug fixes -Prevent filenames in command line programs from using scientific notation when rendering a large number of frames. -Fix tab orders to match newer GUI items which were overlooked in previous releases. -Re-render previews if transparency value in the options dialog was changed. Re-rendering was previously only done if early clip or y axis up was changed. -Use transparency when rendering thumbnail previews. --Code changes -Wrap EmberCommon.h in a namespace called EmberCommon. -Move FormatName() from EmberGenome to EmberCommon.h/cpp -Add a prefix parameter to EmberFile::DefaultFilename() to allow for creating a default filename for sequences. -When showing the final render dialog, allow specifying where it came from: the toolbar or the render sequence button. -Refactor all preview rendering code out into its own class hierarchy with overrides for the main window and the final render dialog. -Remove all preview render cancelling functions, they are now built into the new class hierarchy and a new render will not start until the previous one is stopped. -Add two new function ConstrainLow() and ConstrainHigh() which wrap constraining two min/max spinboxes to each others' values. -Add a bool to FractoriumEmberControllerBase::CopyEmberFile() to specify whether to copy the main file or the sequence file. This is somewhat of a hack and was done in a rush. -Add a bool to FractoriumEmberControllerBase::SetEmberFile() to specify whether to move the file rather than copy. This is used in FinalRenderEmberController and improves efficiency. -Add wrapper functions for variations filter dialog settings.
2016-06-11 20:47:03 -04:00
using namespace EmberCommon;
--User changes: -Show common folder locations such as documents, downloads, pictures in the sidebar in all file dialogs. -Warning message about exceeding memory in final render dialog now suggests strips as the solution to the problem. -Strips now has a tooltip explaining what it does. -Allow more digits in the spinners on the color section the flame tab. -Add manually adjustable size spinners in the final render dialog. Percentage scale and absolute size are fully synced. -Default prefix in final render is now the filename when doing animations (coming from sequence section of the library tab). -Changed the elliptic variation back to using a less precise version for float, and a more precise version for double. The last release had it always using double. -New applied xaos table that shows a read-only view of actual weights by taking the base xform weights and multiplying them by the xaos values. -New table in the xaos tab that gives a graphical representation of the probability that each xform is chosen, with and without xaos. -Add button to transpose the xaos rows and columns. -Add support for importing .chaos files from Chaotica. --Pasting back to Chaotica will work for most, but not all, variations due to incompatible parameter names in some. -Curves are now splines instead of Bezier. This adds compatibility with Chaotica, but breaks it for Apophysis. Xmls are still pastable, but the color curves will look different. --The curve editor on the palette tab can now add points by clicking on the lines and remove points by clicking on the points themselves, just like Chaotica. --Splines are saved in four new xml fields: overall_curve, red_curve, green_curve and blue_curve. -Allow for specifying the percentage of a sub batch each thread should iterate through per kernel call when running with OpenCL. This gives a roughly 1% performance increase due to having to make less kernel calls while iterating. --This field is present for interactive editing (where it's not very useful) and in the final render dialog. --On the command line, this is specified as --sbpctth for EmberRender and EmberAnimate. -Allow double clicking to toggle the supersample field in the flame tab between 1 and 2 for easily checking the effect of the field. -When showing affine values as polar coordinates, show angles normalized to 360 to match Chaotica. -Fuse Count spinner now toggles between 15 and 100 when double clicking for easily checking the effect of the field. -Added field for limiting the range in the x and y direction that the initial points are chosen from. -Added a field called K2 which is an alternative way to set brightness, ignored when zero. --This has no effect for many variations, but hs a noticeable effect for some. -Added new variations: arcsech arcsech2 arcsinh arctanh asteria block bwraps_rand circlecrop2 coth_spiral crackle2 depth_blur depth_blur2 depth_gaussian depth_gaussian2 depth_ngon depth_ngon2 depth_sine depth_sine2 dragonfire dspherical dust excinis exp2 flipx flowerdb foci_p gaussian glynnia2 glynnsim4 glynnsim5 henon henon hex_rand hex_truchet hypershift lazyjess lens lozi lozi modulusx modulusy oscilloscope2 point_symmetry pointsymmetry projective pulse rotate scry2 shift smartshape spher squares starblur2 swirl3 swirl3r tanh_spiral target0 target2 tile_hlp truchet_glyph truchet_inv truchet_knot unicorngaloshen vibration vibration2 --hex_truchet, hex_rand should always use double. They are extremely sensitive. --Bug fixes: -Bounds sign was flipped for x coordinate of world space when center was not zero. -Right clicking and dragging spinner showed menu on mouse up, even if it was very far away. -Text boxes for size in final render dialog were hard to type in. Same bug as xform weight used to be so fix the same way. -Fix spelling to be plural in toggle color speed box. -Stop using the blank user palette to generate flames. Either put colored palettes in it, or exclude it from randoms. -Clicking the random palette button for a palette file with only one palette in it would freeze the program. -Clicking none scale in final render did not re-render the preview. -Use less precision on random xaos. No need for 12 decimal places. -The term sub batch is overloaded in the options dialog. Change the naming and tooltip of those settings for cpu and opencl. --Also made clear in the tooltip for the default opencl quality setting that the value is per device. -The arrows spinner in palette editor appears like a read-only label. Made it look like a spinner. -Fix border colors for various spin boxes and table headers in the style sheet. Requires reload. -Fix a bug in the bwraps variation which would produce different results than Chaotica and Apophysis. -Synth was allowed to be selected for random flame generation when using an Nvidia card but it shouldn't have been because Nvidia has a hard time compiling synth. -A casting bug in the OpenCL kernels for log scaling and density filtering was preventing successful compilations on Intel iGPUs. Fixed even though we don't support anything other than AMD and Nvidia. -Palette rotation (click and drag) position was not being reset when loading a new flame. -When the xform circles were hidden, opening and closing the options dialog would improperly reshow them. -Double click toggle was broken on integer spin boxes. -Fixed tab order of some controls. -Creating a palette from a jpg in the palette editor only produced a single color. --Needed to package imageformats/qjpeg.dll with the Windows installer. -The basic memory benchmark test flame was not really testing memory. Make it more spread out. -Remove the temporal samples field from the flame tab, it was never used because it's only an animation parameter which is specified in the final render dialog or on the command line with EmberAnimate. --Code changes: -Add IsEmpty() to Palette to determine if a palette is all black. -Attempt to avoid selecting a blank palette in PaletteList::GetRandomPalette(). -Add function ScanForChaosNodes() and some associated helper functions in XmlToEmber. -Make variation param name correction be case insensitive in XmlToEmber. -Report error when assigning a variation param value in XmlToEmber. -Add SubBatchPercentPerThread() method to RendererCL. -Override enterEvent() and leaveEvent() in DoubleSpinBox and SpinBox to prevent the context menu from showing up on right mouse up after already leaving the spinner. -Filtering the mouse wheel event in TableWidget no longer appears to be needed. It was probably an old Qt bug that has been fixed. -Gui/ember syncing code in the final render dialog needed to be reworked to accommodate absolute sizes.
2019-04-13 22:00:46 -04:00
#define DO_NVIDIA 1
Features: --Add support for Exr files which use 32-bit floats for each RGBA channel. Seems to come out too washed out. --Allow for clearing an individual color curve. --Allow for saving multiple image types in EmberRender and EmberAnimate. All writes are threaded. --Remove --bpc command line argument. Add format png16 as a replacement. --Remove --enable_jpg_comments and --enable_png_comments command line arguments, and replace them with --enable_comments which applies to jpg, png and exr. --Add menu items to variations and affine spinners which allow for easy entry of specific numeric values like pi. --Make final render dialog be wider rather than so tall. Bug fixes: --Fix some OpenCL compile errors on Mac. --Remove ability to save bitmap files on all platforms but Windows. Code changes: --New dependency on OpenEXR. --Allow Curves class to interact with objects of a different template type. --Make m_Curves member of Ember always use float as template type. --Set the length of the curves array to always be 2^17 which should offer enough precision with new 32-bit float pixel types. --Set pixel types to always be 32-bit float. This results in a major reduction of code in the final accumulation part of Renderer.h/cpp. --Remove corresponding code from RendererCL and FinalAccumOpenCLKernelCreator. --Remove Transparency, NumChannels and BytesPerPixel setters from Renderer.h/cpp. --Add new global functions to format final image buffers and place all alpha calculation and scaling code in them. --Blending is no longer needed in OpenGLWidget because of the new pixel type. --Make new class, AffineDoubleSpinBox. --Attempt to make file save dialog code work the same on all OSes. --Remove some unused functions.
2017-07-22 16:43:35 -04:00
void writeRgba1(const char filename[],
const Rgba* pixels,
int width,
int height)
{
//
// Write an RGBA image using class RgbaOutputFile.
//
// - open the file
// - describe the memory layout of the pixels
// - store the pixels in the file
//
//auto& chl = file.header().channels();
//chl.findChannel("R")->type = PixelType::FLOAT;
//Header header(width, height);
//header.channels().insert("R", Channel(IMF::FLOAT));
//header.channels().insert("G", Channel(IMF::FLOAT));
//header.channels().insert("B", Channel(IMF::FLOAT));
////header.channels().insert("A", Channel(IMF::FLOAT));
//FrameBuffer frameBuffer;
//frameBuffer.insert("Z", // name
// Slice(IMF::FLOAT, // type
// (char *)zPixels, // base
// sizeof(*zPixels) * 1, // xStride
// sizeof(*zPixels) * width)); // yStride
try
{
RgbaOutputFile file(filename, width, height, WRITE_RGBA);
file.setFrameBuffer(pixels, 1, width);
file.writePixels(height);
}
catch (std::exception e)
{
cout << e.what() << endl;
}
}
template <typename T>
void SaveFinalImage(Renderer<T, T>& renderer, vector<byte>& pixels, char* suffix)
{
long newSize;
2016-12-05 22:04:33 -05:00
ostringstream os;
os << ".\\BasicFlame_" << sizeof(T) << "_" << suffix ".bmp";
BYTE* bgrBuf = ConvertRGBToBMPBuffer(pixels.data(), renderer.FinalRasW(), renderer.FinalRasH(), newSize);
SaveBMP(ch, bgrBuf, renderer.FinalRasW(), renderer.FinalRasH(), newSize);
delete [] bgrBuf;
}
template <typename T>
Ember<T> CreateBasicEmber(uint width, uint height, uint ss, T quality, T centerX, T centerY, T rotate)
{
Timing t;
QTIsaac<ISAAC_SIZE, ISAAC_INT> rand;
//t.Tic();
Ember<T> ember1;
//t.Toc("TestBasicFlame() : Constructor()");
//t.Tic();
ember1.m_FinalRasW = width;
ember1.m_FinalRasH = height;
ember1.m_Supersample = ss;
ember1.m_Quality = quality;
ember1.m_CenterX = centerX;
ember1.m_CenterY = centerY;
ember1.m_Rotate = rotate;
Xform<T> xform1(T(0.25), rand.Frand01<T>(), rand.Frand11<T>(), T(1), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
Xform<T> xform2(T(0.25), rand.Frand01<T>(), rand.Frand11<T>(), T(1), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
Xform<T> xform3(T(0.25), rand.Frand01<T>(), rand.Frand11<T>(), T(1), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
Xform<T> xform4(T(0.25), rand.Frand01<T>(), rand.Frand11<T>(), T(1), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
xform1.AddVariation(new SphericalVariation<T>());
xform2.AddVariation(new SphericalVariation<T>());
xform3.AddVariation(new SphericalVariation<T>());
xform4.AddVariation(new SphericalVariation<T>());
xform4.AddVariation(new BlobVariation<T>());
ember1.AddXform(xform1);
ember1.AddXform(xform2);
ember1.AddXform(xform3);
ember1.AddXform(xform4);
//ember1.SetFinalXform(xform4);
return ember1;
}
string GetEmberCLKernelString(Ember<float>& ember, bool iter, bool log, bool de, uint ss, bool accum)
{
ostringstream os;
IterOpenCLKernelCreator<float> iterCreator;
DEOpenCLKernelCreator deCreator(false, false);
FinalAccumOpenCLKernelCreator accumCreator(false);
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
pair<string, vector<float>> pair, pair2;
iterCreator.ParVarIndexDefines(ember, pair);
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
iterCreator.SharedDataIndexDefines(ember, pair2);
if (iter)
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
os << "Iter kernel: \n" << iterCreator.CreateIterKernelString(ember, pair.first, pair2.first, true);
if (log)
os << "Log scale de kernel: \n" << deCreator.LogScaleAssignDEKernel();
//if (de)
// os << "Gaussian DE kernel: \n" << deCreator.GaussianDEKernel(ss);
//if (accum)
// os << "Accum kernel: \n" << accumCreator.FinalAccumKernelLateClipWithoutAlpha();
return os.str();
}
template <typename T>
void MakeTestAllVarsRegPrePost(vector<Ember<T>>& embers)
{
uint index = 0;
ostringstream ss;
auto varList = VariationList<T>::Instance();
auto paletteList = PaletteList<float>::Instance();
QTIsaac<ISAAC_SIZE, ISAAC_INT> rand;
paletteList->Add("flam3-palettes.xml");
Timing t;
Ember<T> emberNoVars;
emberNoVars.m_FinalRasW = 1024;
emberNoVars.m_FinalRasH = 1024;
emberNoVars.m_Quality = 100;
Xform<T> xform1(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
Xform<T> xform2(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
Xform<T> xform3(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
//Xform<T> xform4(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
//Xform<T> xform5(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
//Xform<T> xform6(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
//Xform<T> xform7(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
emberNoVars.AddXform(xform1);
emberNoVars.AddXform(xform2);
emberNoVars.AddXform(xform3);
//emberNoVars.AddXform(xform4);
//emberNoVars.AddXform(xform5);
//emberNoVars.AddXform(xform6);
//emberNoVars.AddXform(xform7);
ss << "NoVars";
emberNoVars.m_Name = ss.str();
ss.str("");
emberNoVars.m_Palette = *paletteList->GetPaletteByFilename(paletteList->m_DefaultFilename, 0);
embers.push_back(emberNoVars);
while (index < varList->RegSize())
{
/* if (index != eVariationId::VAR_SYNTH)
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
{
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
index++;
continue;
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
}
*/
Ember<T> ember1;
unique_ptr<Variation<T>> regVar(varList->GetVariationCopy(index, eVariationType::VARTYPE_REG));
unique_ptr<Variation<T>> preVar(varList->GetVariationCopy("pre_" + regVar->Name()));
unique_ptr<Variation<T>> postVar(varList->GetVariationCopy("post_" + regVar->Name()));
ember1.m_FinalRasW = 1024;
ember1.m_FinalRasH = 1024;
ember1.m_Quality = 500;
Xform<T> xform1(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
Xform<T> xform2(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
Xform<T> xform3(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
Xform<T> xform4(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
Xform<T> xform5(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
Xform<T> xform6(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
Xform<T> xform7(0.25f, rand.Frand01<T>(), rand.Frand11<T>(), 1, rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
if (preVar.get() && postVar.get())
{
xform1.AddVariation(preVar->Copy());
xform2.AddVariation(regVar->Copy());
xform3.AddVariation(postVar->Copy());
xform4.AddVariation(preVar->Copy());
xform4.AddVariation(regVar->Copy());
xform5.AddVariation(preVar->Copy());
xform5.AddVariation(postVar->Copy());
xform6.AddVariation(regVar->Copy());
xform6.AddVariation(postVar->Copy());
xform7.AddVariation(preVar->Copy());
xform7.AddVariation(regVar->Copy());
xform7.AddVariation(postVar->Copy());
ember1.AddXform(xform1);
ember1.AddXform(xform2);
ember1.AddXform(xform3);
ember1.AddXform(xform4);
ember1.AddXform(xform5);
ember1.AddXform(xform6);
ember1.AddXform(xform7);
}
else
{
xform1.AddVariation(regVar->Copy());
xform2.AddVariation(regVar->Copy());
xform3.AddVariation(regVar->Copy());
xform4.AddVariation(regVar->Copy());
ember1.AddXform(xform1);
ember1.AddXform(xform2);
ember1.AddXform(xform3);
ember1.AddXform(xform4);
}
ss << index << "_" << regVar->Name();
ember1.m_Name = ss.str();
ss.str("");
ember1.m_Palette = *paletteList->GetRandomPalette();
index++;
embers.push_back(ember1);
}
t.Toc("Creating embers for all possible variations");
}
void MakeTestAllVarsRegPrePostComboFile(const string& filename)
{
EmberToXml<float> writer;
vector<Ember<float>> embers;
MakeTestAllVarsRegPrePost(embers);
writer.Save(filename, embers, 0, true, true, false, false, false);
}
void TestAtomicAdd()
{
size_t i;
ostringstream os;
OpenCLWrapper wrapper;
vector<float> vec(32);
os << ConstantDefinesString(false) << UnionCLStructString << endl;
os <<
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
"void AtomicAdd(volatile __global float* source, const float operand)\n"
"{\n"
" union\n"
" {\n"
" uint intVal;\n"
" float floatVal;\n"
" } newVal;\n"
"\n"
" union\n"
" {\n"
" uint intVal;\n"
" float floatVal;\n"
" } prevVal;\n"
"\n"
" do\n"
" {\n"
" prevVal.floatVal = *source;\n"
" newVal.floatVal = prevVal.floatVal + operand;\n"
" } while (atomic_cmpxchg((volatile __global uint*)source, prevVal.intVal, newVal.intVal) != prevVal.intVal);\n"
"}\n"
"\n"
"__kernel void MyKernel(\n"
" __global float* buff,\n"
" uint lockit\n"
"\t)\n"
"{\n"
" uint index = THREAD_ID_X;\n"
"\n"
" if (lockit)\n"
" {\n"
" AtomicAdd(&(buff[index]), (float)index * 0.54321);\n"
" }\n"
" else\n"
" {\n"
" buff[index] += (float)index * 0.54321;\n"
" }\n"
"}\n";
string program = os.str();
string entry = "MyKernel";
if (wrapper.Init(0, 0))
{
for (i = 0; i < vec.size(); i++)
vec[i] = (i * 10.2234f);
if (wrapper.AddAndWriteBuffer("buff", (void*)vec.data(), (uint)vec.size() * sizeof(vec[0])))
{
if (wrapper.AddProgram(entry, program, entry, false))
{
wrapper.SetBufferArg(0, 0, 0);
wrapper.SetArg<uint>(0, 1, 0);
if (wrapper.RunKernel(0,
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
32,//Total grid dims.
1,
1,
1,//Individual block dims.
1,
1))
{
wrapper.ReadBuffer(0, vec.data(), (uint)vec.size() * sizeof(vec[0]));
cout << "Vector after unlocked add: " << endl;
for (i = 0; i < vec.size(); i++)
{
cout << "vec[" << i << "] = " << vec[i] << endl;
}
for (i = 0; i < vec.size(); i++)
vec[i] = (i * 10.2234f);
wrapper.AddAndWriteBuffer("buff", (void*)vec.data(), (uint)vec.size() * sizeof(vec[0]));
wrapper.SetBufferArg(0, 0, 0);
wrapper.SetArg<uint>(0, 1, 1);
if (wrapper.RunKernel(0,
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
32,//Total grid dims.
1,
1,
1,//Individual block dims.
1,
1))
{
wrapper.ReadBuffer(0, vec.data(), (uint)vec.size() * sizeof(vec[0]));
cout << "\n\nVector after locked add: " << endl;
for (i = 0; i < vec.size(); i++)
{
cout << "vec[" << i << "] = " << vec[i] << endl;
}
}
}
}
}
}
}
bool TestVarCounts()
{
auto vlf(VariationList<float>::Instance());
#ifdef DO_DOUBLE
auto vld(VariationList<double>::Instance());
bool success((vlf->Size() == vld->Size()) && (vlf->Size() == size_t(eVariationId::LAST_VAR)));
#else
bool success = true;
#endif
uint start = et(eVariationId::VAR_ARCH);
if (!success)
{
cout << "Variation list size " << vlf->Size() << " does not equal the max var ID enum " << et(eVariationId::LAST_VAR) << "." << endl;
}
for (; start < et(eVariationId::LAST_VAR); start++)
{
auto var = vlf->GetVariation((eVariationId)start);
if (!var)
{
cout << "Variation " << start << " was not found." << endl;
success = false;
}
}
return success;
}
template <typename T>
bool TestVarUnique()
{
bool success = true;
auto vl = VariationList<T>::Instance();
vector<eVariationId> ids;
vector<string> names;
ids.reserve(vl->Size());
names.reserve(vl->Size());
for (size_t i = 0; i < vl->Size(); i++)
{
auto var = vl->GetVariation(i);
if (std::find(ids.begin(), ids.end(), var->VariationId()) != ids.end())
{
cout << "Variation " << var->Name() << " was a duplicate ID entry." << endl;
success = false;
}
else
{
ids.push_back(var->VariationId());
}
if (std::find(names.begin(), names.end(), var->Name()) != names.end())
{
cout << "Variation " << var->Name() << " was a duplicate name entry." << endl;
success = false;
}
else
{
names.push_back(var->Name());
}
}
return success;
}
template <typename sT, typename dT>
bool TestVarPrecalcEqual(const Variation<sT>* var1, const Variation<dT>* var2)
{
bool success = true;
if (var1 && var2)
{
if (var1->NeedPrecalcSumSquares() != var2->NeedPrecalcSumSquares())
{
cout << "NeedPrecalcSumSquares value of " << var1->NeedPrecalcSumSquares() << " for variation " << var1->Name() << " != NeedPrecalcSumSquares value of " << var2->NeedPrecalcSumSquares() << " for variation " << var2->Name() << endl;
success = false;
}
if (var1->NeedPrecalcSqrtSumSquares() != var2->NeedPrecalcSqrtSumSquares())
{
cout << "NeedPrecalcSqrtSumSquares value of " << var1->NeedPrecalcSqrtSumSquares() << " for variation " << var1->Name() << " != NeedPrecalcSqrtSumSquares value of " << var2->NeedPrecalcSqrtSumSquares() << " for variation " << var2->Name() << endl;
success = false;
}
if (var1->NeedPrecalcAngles() != var2->NeedPrecalcAngles())
{
cout << "NeedPrecalcAngles value of " << var1->NeedPrecalcAngles() << " for variation " << var1->Name() << " != NeedPrecalcAngles value of " << var2->NeedPrecalcAngles() << " for variation " << var2->Name() << endl;
success = false;
}
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
if (var1->NeedPrecalcAtanXY() != var2->NeedPrecalcAtanXY())
{
cout << "NeedPrecalcAtanXY value of " << var1->NeedPrecalcAtanXY() << " for variation " << var1->Name() << " != NeedPrecalcAtanXY value of " << var2->NeedPrecalcAtanXY() << " for variation " << var2->Name() << endl;
success = false;
}
if (var1->NeedPrecalcAtanYX() != var2->NeedPrecalcAtanYX())
{
cout << "NeedPrecalcAtanYX value of " << var1->NeedPrecalcAtanYX() << " for variation " << var1->Name() << " != NeedPrecalcAtanYX value of " << var2->NeedPrecalcAtanYX() << " for variation " << var2->Name() << endl;
success = false;
}
}
return success;
}
template <typename sT, typename dT>
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
bool TestVarEqual(const Variation<sT>* var1, const Variation<dT>* var2)
{
bool success = true;
if (!var1 || !var2)
{
cout << "Variations were null." << endl;
return false;
}
if (var1->VariationId() != var2->VariationId())
{
cout << "Variation IDs were not equal: " << et(var1->VariationId()) << " != " << et(var2->VariationId()) << endl;
success = false;
}
if (var1->VarType() != var2->VarType())
{
cout << "Variation types were not equal: " << et(var1->VarType()) << " != " << et(var2->VarType()) << endl;
success = false;
}
if (var1->Name() != var2->Name())
{
cout << "Variation names were not equal: " << var1->Name() << " != " << var2->Name() << endl;
success = false;
}
if (var1->Prefix() != var2->Prefix())
{
cout << "Variation prefixes were not equal: " << var1->Prefix() << " != " << var2->Prefix() << endl;
success = false;
}
if (!TestVarPrecalcEqual<sT, dT>(var1, var2))
{
cout << "Variation precalcs were not equal: " << var1->Name() << " and " << var2->Name() << "." << endl;
success = false;
}
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
auto parVar1 = dynamic_cast<const ParametricVariation<sT>*>(var1);
auto parVar2 = dynamic_cast<const ParametricVariation<dT>*>(var2);
if (parVar1 && parVar2)
{
if (parVar1->ParamCount() != parVar2->ParamCount())
{
cout << "Variation ParamCount were not equal: " << parVar1->ParamCount() << " != " << parVar2->ParamCount() << endl;
success = false;
}
vector<ParamWithName<sT>> params1 = parVar1->ParamsVec();
vector<ParamWithName<dT>> params2 = parVar2->ParamsVec();
for (size_t i = 0; i < params1.size(); i++)
{
if (params1[i].Name() != params2[i].Name())
{
cout << "Param Names were not equal: " << params1[i].Name() << " != " << params2[i].Name() << endl;
success = false;
}
if (params1[i].Type() != params2[i].Type())
{
cout << "Param " << params1[i].Name() << " Types were not equal: " << params1[i].Type() << " != " << params2[i].Type() << endl;
success = false;
}
if (params1[i].IsPrecalc() != params2[i].IsPrecalc())
{
cout << "Param " << params1[i].Name() << " IsPrecalc were not equal: " << params1[i].IsPrecalc() << " != " << params2[i].IsPrecalc() << endl;
success = false;
}
if (!IsClose<sT>(params1[i].Def(), (sT)params2[i].Def()))
{
cout << "Param " << params1[i].Name() << " Def were not equal: " << params1[i].Def() << " != " << params2[i].Def() << endl;
success = false;
}
if (typeid(sT) == typeid(dT))//Min and max can be different for float and double.
{
if (!IsClose<sT>(params1[i].Min(), (sT)params2[i].Min()))
{
cout << "Param " << params1[i].Name() << " Min were not equal: " << params1[i].Min() << " != " << params2[i].Min() << endl;
success = false;
}
if (!IsClose<sT>(params1[i].Max(), (sT)params2[i].Max()))
{
cout << "Param " << params1[i].Name() << " Max were not equal: " << params1[i].Max() << " != " << params2[i].Max() << endl;
success = false;
}
}
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
if (!params1[i].IsState() && !params2[i].IsState())//Don't compare state params, they can be different if set to random vals.
{
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
if (!IsClose<sT>(params1[i].ParamVal(), (sT)params2[i].ParamVal(), sT(1e-4)))
{
--User changes -Allow for pausing the renderer in the main window. This makes is more efficient when entering many parameters, such as when following a tutorial. -Add support for new variations: erf, gamma, jac_cn, jac_dn, jac_sn, logDB, pressure_wave, pRose3D, splits3D, w, waves2b, x, xerf, y, z. -Inform user of the start and stop of file parsing in EmberAnimate because the files could potentially be very large. -Move the follwing fields to a new table called Animation: Interpolation, Affine Interpolation, Temporal Samples, Temporal Filter Width, Temporal Filter Type. -These currently have no effect on the interactive renderer and instead are used when running flames through EmberGenome to generate sequences, and then animating them in Fractorium or EmberAnimate. -Add new parameter overrides for EmberRender and EmberAnimate which directly assign values to all flames being rendered, rather than scale: --quality --demin --demax --Bug fixes -Left pad instead of right pad names of sequence outputs from EmberGenome. -Unique file naming was broken for files which already had an underscore in them. -Properly report that png is the default format of EmberRender and EmberAnimate output instead of erroneously claiming it was jpg. -Make command line programs search these folders in this order for the palette file: ./ ~/.fractorium ~/.config/fractorium /usr/share/fractorium /usr/local/share/fractorium -Fix possible bad values in hexes. -Better assignment of Z variables. -Fix boarders due to use of poorly implemented rint() function from flam3. Use std::rint() now. -wedge_sph was completely wrong due to having accidentally swapped the mapping of two parameters. -Make juliascope juliascope_power parameter be of type REAL_NONZERO since it's used as a denominator. -Make Z assignment compatible with the originals in: -arch, bcircle, bCollide, bent, bent2, bisplit, blob, blur_linear, blur_square, bMod, boarders, boarders2, bSwirl, bTransform, butterfly, cardioid, cell, circleblur, circlize, circlize2, circus, collideoscope, cos, cosine, cosh, coth, cpow, cpow2, crescents, cropn, csc, csch, curl, curve, dc_gridout, deltaa, diamond, disc2, eclipse, eCollide, edisc, eJulia, elliptic, eMod, eMotion, ennepers, epispiral, ePush, eRotate, eScale, eSwirl, ex, exp, expo, exponential, fan, fdisc, fibonacci, fibonacci2, fisheye, flipcircle, flipy, flower, flux, funnel, glynnia, GlynnSim1, GlynnSim2, GlynnSim3, gridout, handkerchief, heart, hole, idisc, julia, julian2, juliaNab, kaleidoscope, lazyTravis, Lissajous, mask, MobiusN, mobius_strip, modulus, murl, murl2, npolar, ortho, oscilloscope, parabola, perspective, petal, phoenix_julia, pie (was also inconsistent between cpu and gpu), poincare, popcorn, popcorn2, power, pow_block, rational3, rays, rblur, rings, rippled, roundspher, sec, secant2, sigmoid, sin, sineblur, sinh, sinusgrid, sphericaln, spiralwing, spirograph, split, squarize, squirrel, squish, sschecks, starblur, stripes, stwin, super_shape, tan, tancos, tangent, tanh, TwinTrian, twoface, unpolar, waves, wavesn, wedge_julia, whorl, xheart, zblur, zscale. --Code changes -Generalize Variation::PrecalcHelper() and rename to PrePostPrecalcHelper(). --Do the same for the OpenCL version and rename it PrePostPrecalcOpenCLString(). -Rename Variation::m_AssignType to m_PrePostAssignType since it's only relevant to pre/post variations.
2016-01-29 20:02:15 -05:00
cout << "Param " << params1[i].Name() << " Vals were not equal: " << params1[i].ParamVal() << " != " << params2[i].ParamVal() << endl;
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
success = false;
}
}
}
}
return success;
}
bool TestVarPrePostNames()
{
bool success = true;
auto vlf(VariationList<float>::Instance());
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
string name = var->Name();
if (var->VarType() == eVariationType::VARTYPE_REG)
{
if (name.find("pre_") == 0)
{
cout << "Regular variation " << name << " must not start with pre_." << endl;
success = false;
}
if (name.find("post_") == 0)
{
cout << "Regular variation " << name << " must not start with post_." << endl;
success = false;
}
}
else if (var->VarType() == eVariationType::VARTYPE_PRE)
{
if (name.find("pre_") != 0)
{
cout << "Pre variation " << name << " must start with pre_." << endl;
success = false;
}
}
else if (var->VarType() == eVariationType::VARTYPE_POST)
{
if (name.find("post_") != 0)
{
cout << "Post variation " << name << " must start with post_." << endl;
success = false;
}
}
else
{
cout << "Invalid variation type." << endl;
success = false;
break;
}
if (auto parVar = dynamic_cast<ParametricVariation<float>*>(const_cast<Variation<float>*>(var)))
{
vector<ParamWithName<float>> params = parVar->ParamsVec();
for (size_t p = 0; p < params.size(); p++)
{
if (params[p].Name().find(name.c_str()) != 0)
{
cout << "Param " << params[p].Name() << " must start with " << name << endl;
success = false;
}
}
}
}
return success;
}
template <typename sT, typename dT>
bool TestVarCopy()
{
bool success = true;
auto vlf(VariationList<sT>::Instance());
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
Variation<dT>* destVar = NULL;
unique_ptr<Variation<sT>> copyVar(var->Copy());//Test Copy().
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
if (!TestVarEqual<sT, sT>(var, copyVar.get()))
{
cout << "Variations " << var->Name() << "<" << typeid(sT).name() << "> and " << copyVar->Name() << "<" << typeid(sT).name() << "> (same template type) were not equal after Copy()." << endl;
success = false;
}
var->Copy(destVar);//Test Copy(var*);
unique_ptr<Variation<dT>> destPtr(destVar);//Just for deletion.
if (!TestVarEqual<sT, dT>(var, destPtr.get()))
{
cout << "Variations " << var->Name() << "<" << typeid(sT).name() << "> and " << destPtr->Name() << "<" << typeid(dT).name() << "> (different template types) were not equal after Copy(Variation<T>*)." << endl;
success = false;
}
}
return success;
}
bool TestParVars()
{
bool success = true;
auto vlf(VariationList<float>::Instance());
for (size_t i = 0; i < vlf->ParametricSize(); i++)
{
if (auto parVar = vlf->GetParametricVariation(i))
{
if (parVar->ParamCount() < 1)
{
cout << "Parametric variation " << parVar->Name() << " does not have any parameters." << endl;
success = false;
}
vector<string> names;
vector<float*> addresses;
auto params = parVar->Params();
names.reserve(parVar->ParamCount());
addresses.reserve(parVar->ParamCount());
--User changes -Add support for multiple GPU devices. --These options are present in the command line and in Fractorium. -Change scheme of specifying devices from platform,device to just total device index. --Single number on the command line. --Change from combo boxes for device selection to a table of all devices in Fractorium. -Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Bug fixes -EmberAnimate, EmberRender, FractoriumSettings, FinalRenderDialog: Fix wrong order of arguments to Clamp() when assigning thread priority. -VariationsDC.h: Fix NVidia OpenCL compilation error in DCTriangleVariation. -FractoriumXformsColor.cpp: Checking for null pixmap pointer is not enough, must also check if the underlying buffer is null via call to QPixmap::isNull(). --Code changes -Ember.h: Add case for FLAME_MOTION_NONE and default in ApplyFlameMotion(). -EmberMotion.h: Call base constructor. -EmberPch.h: #pragma once only on Windows. -EmberToXml.h: --Handle different types of exceptions. --Add default cases to ToString(). -Isaac.h: Remove unused variable in constructor. -Point.h: Call base constructor in Color(). -Renderer.h/cpp: --Add bool to Alloc() to only allocate memory for the histogram. Needed for multi-GPU. --Make CoordMap() return a const ref, not a pointer. -SheepTools.h: --Use 64-bit types like the rest of the code already does. --Fix some comment misspellings. -Timing.h: Make BeginTime(), EndTime(), ElapsedTime() and Format() be const functions. -Utils.h: --Add new functions Equal() and Split(). --Handle more exception types in ReadFile(). --Get rid of most legacy blending of C and C++ argument parsing. -XmlToEmber.h: --Get rid of most legacy blending of C and C++ code from flam3. --Remove some unused variables. -EmberAnimate: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --If a render fails, exit since there is no point in continuing an animation with a missing frame. --Pass variables to threaded save better, which most likely fixes a very subtle bug that existed before. --Remove some unused variables. -EmberGenome, EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. -EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --Only print values when not rendering with OpenCL, since they're always 0 in that case. -EmberCLPch.h: --#pragma once only on Windows. --#include <atomic>. -IterOpenCLKernelCreator.h: Add new kernel for summing two histograms. This is needed for multi-GPU. -OpenCLWrapper.h: --Move all OpenCL info related code into its own class OpenCLInfo. --Add members to cache the values of global memory size and max allocation size. -RendererCL.h/cpp: --Redesign to accomodate multi-GPU. --Constructor now takes a vector of devices. --Remove DumpErrorReport() function, it's handled in the base. --ClearBuffer(), ReadPoints(), WritePoints(), ReadHist() and WriteHist() now optionally take a device index as a parameter. --MakeDmap() override and m_DmapCL member removed because it no longer applies since the histogram is always float since the last commit. --Add new function SumDeviceHist() to sum histograms from two devices by first copying to a temporary on the host, then a temporary on the device, then summing. --m_Calls member removed, as it's now per-device. --OpenCLWrapper removed. --m_Seeds member is now a vector of vector of seeds, to accomodate a separate and different array of seeds for each device. --Added member m_Devices, a vector of unique_ptr of RendererCLDevice. -EmberCommon.h --Added Devices() function to convert from a vector of device indices to a vector of platform,device indices. --Changed CreateRenderer() to accept a vector of devices to create a single RendererCL which will split work across multiple devices. --Added CreateRenderers() function to accept a vector of devices to create multiple RendererCL, each which will render on a single device. --Add more comments to some existing functions. -EmberCommonPch.h: #pragma once only on Windows. -EmberOptions.h: --Remove --platform option, it's just sequential device number now with the --device option. --Make --out be OPT_USE_RENDER instead of OPT_RENDER_ANIM since it's an error condition when animating. It makes no sense to write all frames to a single image. --Add Devices() function to parse comma separated --device option string and return a vector of device indices. --Make int and uint types be 64-bit, so intmax_t and size_t. --Make better use of macros. -JpegUtils.h: Make string parameters to WriteJpeg() and WritePng() be const ref. -All project files: Turn off buffer security check option in Visual Studio (/Gs-) -deployment.pri: Remove the line OTHER_FILES +=, it's pointless and was causing problems. -Ember.pro, EmberCL.pro: Add CONFIG += plugin, otherwise it wouldn't link. -EmberCL.pro: Add new files for multi-GPU support. -build_all.sh: use -j4 and QMAKE=${QMAKE:/usr/bin/qmake} -shared_settings.pri: -Add version string. -Remove old DESTDIR definitions. -Add the following lines or else nothing would build: CONFIG(release, debug|release) { CONFIG += warn_off DESTDIR = ../../../Bin/release } CONFIG(debug, debug|release) { DESTDIR = ../../../Bin/debug } QMAKE_POST_LINK += $$quote(cp --update ../../../Data/flam3-palettes.xml $${DESTDIR}$$escape_expand(\n\t)) LIBS += -L/usr/lib -lpthread -AboutDialog.ui: Another futile attempt to make it look correct on Linux. -FinalRenderDialog.h/cpp: --Add support for multi-GPU. --Change from combo boxes for device selection to a table of all devices. --Ensure device selection makes sense. --Remove "FinalRender" prefix of various function names, it's implied given the context. -FinalRenderEmberController.h/cpp: --Add support for multi-GPU. --Change m_FinishedImageCount to be atomic. --Move CancelRender() from the base to FinalRenderEmberController<T>. --Refactor RenderComplete() to omit any progress related functionality or image saving since it can be potentially ran in a thread. --Consolidate setting various renderer fields into SyncGuiToRenderer(). -Fractorium.cpp: Allow for resizing of the options dialog to show the entire device table. -FractoriumCommon.h: Add various functions to handle a table showing the available OpenCL devices on the system. -FractoriumEmberController.h/cpp: Remove m_FinalImageIndex, it's no longer needed. -FractoriumRender.cpp: Scale the interactive sub batch count and quality by the number of devices used. -FractoriumSettings.h/cpp: --Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Add multi-GPU support, remove old device,platform pair. -FractoriumToolbar.cpp: Disable OpenCL toolbar button if there are no devices present on the system. -FractoriumOptionsDialog.h/cpp: --Add support for multi-GPU. --Consolidate more assignments in DataToGui(). --Enable/disable CPU/OpenCL items in response to OpenCL checkbox event. -Misc: Convert almost everything to size_t for unsigned, intmax_t for signed.
2015-09-12 21:33:45 -04:00
for (size_t j = 0; j < parVar->ParamCount(); j++)
{
if (std::find(names.begin(), names.end(), params[j].Name()) != names.end())
{
cout << "Param " << params[j].Name() << " for variation " << parVar->Name() << " was a duplicate name entry." << endl;
success = false;
}
else
{
names.push_back(params[j].Name());
}
if (std::find(addresses.begin(), addresses.end(), params[j].Param()) != addresses.end())
{
cout << "Param address" << params[j].Param() << " for variation " << parVar->Name() << " was a duplicate name entry." << endl;
success = false;
}
else
{
addresses.push_back(params[j].Param());
}
}
}
else
{
cout << "Parametric variation at index " << i << " was NULL." << endl;
success = false;
}
}
return success;
}
bool TestVarRegPrePost()
{
bool success = true;
auto vlf(VariationList<float>::Instance());
for (size_t i = 0; i < vlf->RegSize(); i++)
{
auto regVar = vlf->GetVariation(i, eVariationType::VARTYPE_REG);
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
if (regVar)
{
if (regVar->Name().find("dc_") != 0)
{
string name = regVar->Name();
auto preVar = vlf->GetVariation("pre_" + name);
auto postVar = vlf->GetVariation("post_" + name);
if (!preVar)
{
cout << "Pre variation equivalent of " << name << " could not be found." << endl;
success = false;
}
if (!postVar)
{
cout << "Post variation equivalent of " << name << " could not be found." << endl;
success = false;
}
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
if (!TestVarPrecalcEqual<float, float>(regVar, preVar))
{
cout << "Regular and pre variation precalc test failed for " << regVar->Name() << " and " << preVar->Name() << "." << endl;
success = false;
}
if (!TestVarPrecalcEqual<float, float>(regVar, postVar))
{
cout << "Regular and post variation precalc test failed for " << regVar->Name() << " and " << postVar->Name() << "." << endl;
success = false;
}
}
}
else
{
cout << "Regular variation " << i << " was NULL." << endl;
success = false;
}
}
return success;
}
bool TestVarPrecalcUsedCL()
{
bool success = true;
auto vlf(VariationList<float>::Instance());
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
string s = var->OpenCLString();
if (var->NeedPrecalcAngles())
{
if (s.find("precalcSina") == string::npos)
{
cout << "Variation " << var->Name() << " needed precalcSina, but it wasn't found in the OpenCL string." << endl;
success = false;
}
if (s.find("precalcCosa") == string::npos)
{
cout << "Variation " << var->Name() << " needed precalcCosa, but it wasn't found in the OpenCL string." << endl;
success = false;
}
}
else
{
if (s.find("precalcSina") != string::npos)
{
cout << "Variation " << var->Name() << " didn't need precalcSina, but it was found in the OpenCL string." << endl;
success = false;
}
if (s.find("precalcCosa") != string::npos)
{
cout << "Variation " << var->Name() << " didn't need precalcCosa, but it was found in the OpenCL string." << endl;
success = false;
}
if (var->NeedPrecalcSqrtSumSquares())
{
if (s.find("precalcSqrtSumSquares") == string::npos)
{
cout << "Variation " << var->Name() << " needed precalcSqrtSumSquares, but it wasn't found in the OpenCL string." << endl;
success = false;
}
}
else
{
if (s.find("precalcSqrtSumSquares") != string::npos)
{
cout << "Variation " << var->Name() << " didn't need precalcSqrtSumSquares, but it was found in the OpenCL string." << endl;
success = false;
}
if (var->NeedPrecalcSumSquares())
{
if (s.find("precalcSumSquares") == string::npos)
{
cout << "Variation " << var->Name() << " needed precalcSumSquares, but it wasn't found in the OpenCL string." << endl;
success = false;
}
}
else
{
if (s.find("precalcSumSquares") != string::npos)
{
cout << "Variation " << var->Name() << " didn't need precalcSumSquares, but it was found in the OpenCL string." << endl;
success = false;
}
}
}
}
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
if (var->NeedPrecalcSumSquares())
{
if (s.find("SQR(vIn.x) + SQR(vIn.y)") != string::npos || s.find("vIn.x * vIn.x + vIn.y * vIn.y") != string::npos)
{
cout << "Variation " << var->Name() << " needed precalcSumSquares, but is not using it properly." << endl;
success = false;
}
}
else
{
if (s.find("precalcSumSquares") != string::npos)
{
cout << "Variation " << var->Name() << " didn't need precalcSumSquares, but it was found in the OpenCL string." << endl;
success = false;
}
if (s.find("SQR(vIn.x) + SQR(vIn.y)") != string::npos || s.find("vIn.x * vIn.x + vIn.y * vIn.y") != string::npos)
{
cout << "Variation " << var->Name() << " did not specify precalcSumSquares, but could benefit from it." << endl;
success = false;
}
}
if (var->NeedPrecalcSqrtSumSquares())
{
if (s.find("sqrt(SQR(vIn.x) + SQR(vIn.y))") != string::npos || s.find("sqrt(vIn.x * vIn.x + vIn.y * vIn.y)") != string::npos)
{
cout << "Variation " << var->Name() << " needed precalcSqrtSumSquares, but is not using it properly." << endl;
success = false;
}
}
else
{
if (s.find("precalcSqrtSumSquares") != string::npos)
{
cout << "Variation " << var->Name() << " didn't need precalcSqrtSumSquares, but it was found in the OpenCL string." << endl;
success = false;
}
if (s.find("sqrt(SQR(vIn.x) + SQR(vIn.y))") != string::npos || s.find("sqrt(vIn.x * vIn.x + vIn.y * vIn.y)") != string::npos)
{
cout << "Variation " << var->Name() << " did not specify precalcSqrtSumSquares, but could benefit from it." << endl;
success = false;
}
}
if (var->NeedPrecalcAtanXY())
{
if (s.find("precalcAtanxy") == string::npos)
{
cout << "Variation " << var->Name() << " needed precalcAtanxy, but it wasn't found in the OpenCL string." << endl;
success = false;
}
}
else
{
if (s.find("precalcAtanxy") != string::npos)
{
cout << "Variation " << var->Name() << " didn't need precalcAtanxy, but it was found in the OpenCL string." << endl;
success = false;
}
if (s.find("atan2(vIn.x, vIn.y)") != string::npos)
{
cout << "Variation " << var->Name() << " did not specify precalcAtanxy, but could benefit from it." << endl;
success = false;
}
}
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
if (var->NeedPrecalcAtanYX())
{
if (s.find("precalcAtanyx") == string::npos)
{
cout << "Variation " << var->Name() << " needed precalcAtanyx, but it wasn't found in the OpenCL string." << endl;
success = false;
}
}
else
{
if (s.find("precalcAtanyx") != string::npos)
{
cout << "Variation " << var->Name() << " didn't need precalcAtanyx, but it was found in the OpenCL string." << endl;
success = false;
}
if (s.find("atan2(vIn.y, vIn.x)") != string::npos)
{
cout << "Variation " << var->Name() << " did not specify precalcAtanyx, but could benefit from it." << endl;
success = false;
}
}
}
return success;
}
bool TestVarAssignTypes()
{
bool success = true;
auto vlf(VariationList<float>::Instance());
vector<string> vset, vsum;
vset.push_back("vIn.x");
vset.push_back("vIn.y");
vset.push_back("vIn.z");
vset.push_back("precalcSumSquares");
vset.push_back("precalcSqrtSumSquares");
vset.push_back("precalcSina");
vset.push_back("precalcCosa");
vset.push_back("precalcAtanxy");
vset.push_back("precalcAtanyx");
vsum.push_back("vIn.x");
vsum.push_back("vIn.y");
//vsum.push_back("vIn.z");
vsum.push_back("precalcSumSquares");
vsum.push_back("precalcSqrtSumSquares");
vsum.push_back("precalcSina");
vsum.push_back("precalcCosa");
vsum.push_back("precalcAtanxy");
vsum.push_back("precalcAtanyx");
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
string s = var->OpenCLString();
//Only test pre and post. The assign type for regular is ignored, and will always be summed.
if (var->VarType() != eVariationType::VARTYPE_REG)
{
if (var->AssignType() == eVariationAssignType::ASSIGNTYPE_SET)
{
if (!SearchVar(var, vset, false))
{
cout << "Variation " << var->Name() << " had an assign type of SET, but did not use its input points. It should have an assign type of SUM." << endl;
success = false;
}
}
else if (var->AssignType() == eVariationAssignType::ASSIGNTYPE_SUM)
{
if (SearchVar(var, vsum, false))
{
cout << "Variation " << var->Name() << " had an assign type of SUM, but used its input points. It should have an assign type of SET." << endl;
success = false;
}
}
else
{
cout << "Variation " << var->Name() << " had an invalid assign type of " << et(var->AssignType()) << endl;
}
}
}
return success;
}
bool TestVarAssignVals()
{
bool success = true;
auto vlf(VariationList<float>::Instance());
vector<string> xout, yout, zout;
xout.push_back("vOut.x =");
xout.push_back("vOut.x +=");
xout.push_back("vOut.x -=");
xout.push_back("vOut.x *=");
xout.push_back("vOut.x /=");
yout.push_back("vOut.y =");
yout.push_back("vOut.y +=");
yout.push_back("vOut.y -=");
yout.push_back("vOut.y *=");
yout.push_back("vOut.y /=");
zout.push_back("vOut.z =");
zout.push_back("vOut.z +=");
zout.push_back("vOut.z -=");
zout.push_back("vOut.z *=");
zout.push_back("vOut.z /=");
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
if (!SearchVar(var, xout, false))
{
cout << "Variation " << var->Name() << " did not set its x output point. If unused, at least pass through or set to 0." << endl;
success = false;
}
if (!SearchVar(var, yout, false))
{
cout << "Variation " << var->Name() << " did not set its y output point. If unused, at least pass through or set to 0." << endl;
success = false;
}
if (!SearchVar(var, zout, false))
{
cout << "Variation " << var->Name() << " did not set its z output point. If unused, at least pass through or set to 0." << endl;
success = false;
}
}
return success;
}
void FindFmaCandidates()
{
auto vlf(VariationList<float>::Instance());
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
auto cl = var->OpenCLFuncsString() + "\n" + var->StateInitOpenCLString() + "\n" + var->OpenCLString();
auto splits = Split(cl, '\n');
for (auto& split : splits)
{
if ((split.find("*") != string::npos || split.find("SQR(") != string::npos || split.find("Sqr(") != string::npos) && split.find("+") != string::npos)
cout << "Variation " << var->Name() << " is a potential fma() candidate because of line:\n\t" << split << endl;
}
}
}
void FindFmaImplemented()
{
auto vlf(VariationList<float>::Instance());
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
auto cl = var->OpenCLFuncsString() + "\n" + var->StateInitOpenCLString() + "\n" + var->OpenCLString();
auto splits = Split(cl, '\n');
for (auto& split : splits)
{
if ((split.find("fma(") != string::npos))
cout << "Variation " << var->Name() << " has been implemented with fma():\n\t" << split << endl;
}
}
}
bool TestZepsFloor()
{
bool success = true;
auto vlf(VariationList<float>::Instance());
vector<string> zeps;
zeps.push_back("Zeps(floor");
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
if (SearchVar(var, zeps, false))
{
cout << "Variation " << var->Name() << " contains Zeps(floor()). This is fine for the GPU, but ensure the CPU uses Zeps<T>((T)Floor<T>())." << endl;
success = false;
}
}
return success;
}
bool TestConstants()
{
bool success = true;
auto vlf(VariationList<float>::Instance());
vector<string> stringVec;
stringVec.push_back("2 * M_PI");
stringVec.push_back("2*M_PI");
stringVec.push_back("M_PI*2");
stringVec.push_back("M_PI * 2");
--User changes: -Show common folder locations such as documents, downloads, pictures in the sidebar in all file dialogs. -Warning message about exceeding memory in final render dialog now suggests strips as the solution to the problem. -Strips now has a tooltip explaining what it does. -Allow more digits in the spinners on the color section the flame tab. -Add manually adjustable size spinners in the final render dialog. Percentage scale and absolute size are fully synced. -Default prefix in final render is now the filename when doing animations (coming from sequence section of the library tab). -Changed the elliptic variation back to using a less precise version for float, and a more precise version for double. The last release had it always using double. -New applied xaos table that shows a read-only view of actual weights by taking the base xform weights and multiplying them by the xaos values. -New table in the xaos tab that gives a graphical representation of the probability that each xform is chosen, with and without xaos. -Add button to transpose the xaos rows and columns. -Add support for importing .chaos files from Chaotica. --Pasting back to Chaotica will work for most, but not all, variations due to incompatible parameter names in some. -Curves are now splines instead of Bezier. This adds compatibility with Chaotica, but breaks it for Apophysis. Xmls are still pastable, but the color curves will look different. --The curve editor on the palette tab can now add points by clicking on the lines and remove points by clicking on the points themselves, just like Chaotica. --Splines are saved in four new xml fields: overall_curve, red_curve, green_curve and blue_curve. -Allow for specifying the percentage of a sub batch each thread should iterate through per kernel call when running with OpenCL. This gives a roughly 1% performance increase due to having to make less kernel calls while iterating. --This field is present for interactive editing (where it's not very useful) and in the final render dialog. --On the command line, this is specified as --sbpctth for EmberRender and EmberAnimate. -Allow double clicking to toggle the supersample field in the flame tab between 1 and 2 for easily checking the effect of the field. -When showing affine values as polar coordinates, show angles normalized to 360 to match Chaotica. -Fuse Count spinner now toggles between 15 and 100 when double clicking for easily checking the effect of the field. -Added field for limiting the range in the x and y direction that the initial points are chosen from. -Added a field called K2 which is an alternative way to set brightness, ignored when zero. --This has no effect for many variations, but hs a noticeable effect for some. -Added new variations: arcsech arcsech2 arcsinh arctanh asteria block bwraps_rand circlecrop2 coth_spiral crackle2 depth_blur depth_blur2 depth_gaussian depth_gaussian2 depth_ngon depth_ngon2 depth_sine depth_sine2 dragonfire dspherical dust excinis exp2 flipx flowerdb foci_p gaussian glynnia2 glynnsim4 glynnsim5 henon henon hex_rand hex_truchet hypershift lazyjess lens lozi lozi modulusx modulusy oscilloscope2 point_symmetry pointsymmetry projective pulse rotate scry2 shift smartshape spher squares starblur2 swirl3 swirl3r tanh_spiral target0 target2 tile_hlp truchet_glyph truchet_inv truchet_knot unicorngaloshen vibration vibration2 --hex_truchet, hex_rand should always use double. They are extremely sensitive. --Bug fixes: -Bounds sign was flipped for x coordinate of world space when center was not zero. -Right clicking and dragging spinner showed menu on mouse up, even if it was very far away. -Text boxes for size in final render dialog were hard to type in. Same bug as xform weight used to be so fix the same way. -Fix spelling to be plural in toggle color speed box. -Stop using the blank user palette to generate flames. Either put colored palettes in it, or exclude it from randoms. -Clicking the random palette button for a palette file with only one palette in it would freeze the program. -Clicking none scale in final render did not re-render the preview. -Use less precision on random xaos. No need for 12 decimal places. -The term sub batch is overloaded in the options dialog. Change the naming and tooltip of those settings for cpu and opencl. --Also made clear in the tooltip for the default opencl quality setting that the value is per device. -The arrows spinner in palette editor appears like a read-only label. Made it look like a spinner. -Fix border colors for various spin boxes and table headers in the style sheet. Requires reload. -Fix a bug in the bwraps variation which would produce different results than Chaotica and Apophysis. -Synth was allowed to be selected for random flame generation when using an Nvidia card but it shouldn't have been because Nvidia has a hard time compiling synth. -A casting bug in the OpenCL kernels for log scaling and density filtering was preventing successful compilations on Intel iGPUs. Fixed even though we don't support anything other than AMD and Nvidia. -Palette rotation (click and drag) position was not being reset when loading a new flame. -When the xform circles were hidden, opening and closing the options dialog would improperly reshow them. -Double click toggle was broken on integer spin boxes. -Fixed tab order of some controls. -Creating a palette from a jpg in the palette editor only produced a single color. --Needed to package imageformats/qjpeg.dll with the Windows installer. -The basic memory benchmark test flame was not really testing memory. Make it more spread out. -Remove the temporal samples field from the flame tab, it was never used because it's only an animation parameter which is specified in the final render dialog or on the command line with EmberAnimate. --Code changes: -Add IsEmpty() to Palette to determine if a palette is all black. -Attempt to avoid selecting a blank palette in PaletteList::GetRandomPalette(). -Add function ScanForChaosNodes() and some associated helper functions in XmlToEmber. -Make variation param name correction be case insensitive in XmlToEmber. -Report error when assigning a variation param value in XmlToEmber. -Add SubBatchPercentPerThread() method to RendererCL. -Override enterEvent() and leaveEvent() in DoubleSpinBox and SpinBox to prevent the context menu from showing up on right mouse up after already leaving the spinner. -Filtering the mouse wheel event in TableWidget no longer appears to be needed. It was probably an old Qt bug that has been fixed. -Gui/ember syncing code in the final render dialog needed to be reworked to accommodate absolute sizes.
2019-04-13 22:00:46 -04:00
stringVec.push_back("2 * MPI");
stringVec.push_back("2*MPI");
stringVec.push_back("MPI*2");
stringVec.push_back("MPI * 2");
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
if (SearchVar(var, stringVec, false))
{
cout << "Variation " << var->Name() << " should be using M_2PI." << endl;
success = false;
--User changes: -Show common folder locations such as documents, downloads, pictures in the sidebar in all file dialogs. -Warning message about exceeding memory in final render dialog now suggests strips as the solution to the problem. -Strips now has a tooltip explaining what it does. -Allow more digits in the spinners on the color section the flame tab. -Add manually adjustable size spinners in the final render dialog. Percentage scale and absolute size are fully synced. -Default prefix in final render is now the filename when doing animations (coming from sequence section of the library tab). -Changed the elliptic variation back to using a less precise version for float, and a more precise version for double. The last release had it always using double. -New applied xaos table that shows a read-only view of actual weights by taking the base xform weights and multiplying them by the xaos values. -New table in the xaos tab that gives a graphical representation of the probability that each xform is chosen, with and without xaos. -Add button to transpose the xaos rows and columns. -Add support for importing .chaos files from Chaotica. --Pasting back to Chaotica will work for most, but not all, variations due to incompatible parameter names in some. -Curves are now splines instead of Bezier. This adds compatibility with Chaotica, but breaks it for Apophysis. Xmls are still pastable, but the color curves will look different. --The curve editor on the palette tab can now add points by clicking on the lines and remove points by clicking on the points themselves, just like Chaotica. --Splines are saved in four new xml fields: overall_curve, red_curve, green_curve and blue_curve. -Allow for specifying the percentage of a sub batch each thread should iterate through per kernel call when running with OpenCL. This gives a roughly 1% performance increase due to having to make less kernel calls while iterating. --This field is present for interactive editing (where it's not very useful) and in the final render dialog. --On the command line, this is specified as --sbpctth for EmberRender and EmberAnimate. -Allow double clicking to toggle the supersample field in the flame tab between 1 and 2 for easily checking the effect of the field. -When showing affine values as polar coordinates, show angles normalized to 360 to match Chaotica. -Fuse Count spinner now toggles between 15 and 100 when double clicking for easily checking the effect of the field. -Added field for limiting the range in the x and y direction that the initial points are chosen from. -Added a field called K2 which is an alternative way to set brightness, ignored when zero. --This has no effect for many variations, but hs a noticeable effect for some. -Added new variations: arcsech arcsech2 arcsinh arctanh asteria block bwraps_rand circlecrop2 coth_spiral crackle2 depth_blur depth_blur2 depth_gaussian depth_gaussian2 depth_ngon depth_ngon2 depth_sine depth_sine2 dragonfire dspherical dust excinis exp2 flipx flowerdb foci_p gaussian glynnia2 glynnsim4 glynnsim5 henon henon hex_rand hex_truchet hypershift lazyjess lens lozi lozi modulusx modulusy oscilloscope2 point_symmetry pointsymmetry projective pulse rotate scry2 shift smartshape spher squares starblur2 swirl3 swirl3r tanh_spiral target0 target2 tile_hlp truchet_glyph truchet_inv truchet_knot unicorngaloshen vibration vibration2 --hex_truchet, hex_rand should always use double. They are extremely sensitive. --Bug fixes: -Bounds sign was flipped for x coordinate of world space when center was not zero. -Right clicking and dragging spinner showed menu on mouse up, even if it was very far away. -Text boxes for size in final render dialog were hard to type in. Same bug as xform weight used to be so fix the same way. -Fix spelling to be plural in toggle color speed box. -Stop using the blank user palette to generate flames. Either put colored palettes in it, or exclude it from randoms. -Clicking the random palette button for a palette file with only one palette in it would freeze the program. -Clicking none scale in final render did not re-render the preview. -Use less precision on random xaos. No need for 12 decimal places. -The term sub batch is overloaded in the options dialog. Change the naming and tooltip of those settings for cpu and opencl. --Also made clear in the tooltip for the default opencl quality setting that the value is per device. -The arrows spinner in palette editor appears like a read-only label. Made it look like a spinner. -Fix border colors for various spin boxes and table headers in the style sheet. Requires reload. -Fix a bug in the bwraps variation which would produce different results than Chaotica and Apophysis. -Synth was allowed to be selected for random flame generation when using an Nvidia card but it shouldn't have been because Nvidia has a hard time compiling synth. -A casting bug in the OpenCL kernels for log scaling and density filtering was preventing successful compilations on Intel iGPUs. Fixed even though we don't support anything other than AMD and Nvidia. -Palette rotation (click and drag) position was not being reset when loading a new flame. -When the xform circles were hidden, opening and closing the options dialog would improperly reshow them. -Double click toggle was broken on integer spin boxes. -Fixed tab order of some controls. -Creating a palette from a jpg in the palette editor only produced a single color. --Needed to package imageformats/qjpeg.dll with the Windows installer. -The basic memory benchmark test flame was not really testing memory. Make it more spread out. -Remove the temporal samples field from the flame tab, it was never used because it's only an animation parameter which is specified in the final render dialog or on the command line with EmberAnimate. --Code changes: -Add IsEmpty() to Palette to determine if a palette is all black. -Attempt to avoid selecting a blank palette in PaletteList::GetRandomPalette(). -Add function ScanForChaosNodes() and some associated helper functions in XmlToEmber. -Make variation param name correction be case insensitive in XmlToEmber. -Report error when assigning a variation param value in XmlToEmber. -Add SubBatchPercentPerThread() method to RendererCL. -Override enterEvent() and leaveEvent() in DoubleSpinBox and SpinBox to prevent the context menu from showing up on right mouse up after already leaving the spinner. -Filtering the mouse wheel event in TableWidget no longer appears to be needed. It was probably an old Qt bug that has been fixed. -Gui/ember syncing code in the final render dialog needed to be reworked to accommodate absolute sizes.
2019-04-13 22:00:46 -04:00
}
}
stringVec.clear();
stringVec.push_back("3 * M_PI");
stringVec.push_back("3*M_PI");
stringVec.push_back("M_PI*3");
stringVec.push_back("M_PI * 3");
//stringVec.push_back("3 * MPI");//Gives a false positive for prose3d.
stringVec.push_back("3*MPI");
stringVec.push_back("MPI*3");
stringVec.push_back("MPI * 3");
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
if (SearchVar(var, stringVec, false))
{
cout << "Variation " << var->Name() << " should be using M_3PI." << endl;
success = false;
}
}
stringVec.clear();
stringVec.push_back("M_PI");
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
if (SearchVar(var, stringVec, false))
{
cout << "Variation " << var->Name() << " is using M_PI, but should be using MPI." << endl;
success = false;
}
}
stringVec.clear();
stringVec.push_back("M_2_PI");
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
if (SearchVar(var, stringVec, false))
{
cout << "Variation " << var->Name() << " is using M_2_PI, but should be using M2PI." << endl;
success = false;
}
}
stringVec.clear();
stringVec.push_back("M_1_PI");
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
if (SearchVar(var, stringVec, false))
{
cout << "Variation " << var->Name() << " is using M_1_PI, but should be using M1PI." << endl;
success = false;
}
}
stringVec.clear();
stringVec.push_back("M_PI_4");
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
if (SearchVar(var, stringVec, false))
{
cout << "Variation " << var->Name() << " is using M_PI_4, but should be using MPI4." << endl;
success = false;
}
}
stringVec.clear();
stringVec.push_back("M_PI_2");
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
if (SearchVar(var, stringVec, false))
{
cout << "Variation " << var->Name() << " is using M_PI_2, but should be using MPI2." << endl;
success = false;
}
}
return success;
}
void TestFuncs()
{
2017-07-01 12:18:01 -04:00
auto vlf(VariationList<float>::Instance());
vector<string> stringVec;
stringVec.push_back("M_E");
stringVec.push_back("M_LOG2E");
stringVec.push_back("M_LOG10E");
stringVec.push_back("M_LN2");
stringVec.push_back("M_LN10");
2017-07-01 12:18:01 -04:00
stringVec.push_back("M_PI");
stringVec.push_back("M_PI_2");
stringVec.push_back("M_PI_4");
stringVec.push_back("M_1_PI");
stringVec.push_back("M_2_PI");
stringVec.push_back("M_2_SQRTPI");
stringVec.push_back("M_SQRT2");
stringVec.push_back("M_SQRT1_2");
//stringVec.push_back("M_2PI");
//stringVec.push_back("M_3PI");
//stringVec.push_back("DEG_2_RAD");
2017-07-01 12:18:01 -04:00
for (size_t i = 0; i < vlf->Size(); i++)
{
auto var = vlf->GetVariation(i);
if (SearchVar(var, stringVec, false))
{
cout << var->Name() << endl;
}
}
}
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
bool TestGlobalFuncs()
{
bool success = true;
auto vlf(VariationList<float>::Instance());
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
vector<string> funcs;
FunctionMapper mapper;
auto funcmap = mapper.GetGlobalMapCopy();
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
for (size_t i = 0; i < vlf->Size(); i++)
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
{
auto var = vlf->GetVariation(i);
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
funcs = var->OpenCLGlobalFuncNames();
auto localfuncs = var->OpenCLFuncsString();
Ember<float> ember;
Xform<float> xf;
xf.AddVariation(var->Copy());
ember.AddXform(xf);
auto kernel = GetEmberCLKernelString(ember, true, false, false, 1u, false);
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
for (auto& func : funcs)//Test if the functions the variation says it requires actually exist.
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
{
if (!mapper.GetGlobalFunc(func))
{
cout << "Variation " << var->Name() << " used unknown global funcion " << func << endl;
success = false;
}
else
{
//cout << "Variation " << var->Name() << " used valid global funcion " << func << endl;
}
}
auto str = var->OpenCLString();
auto vec = var->OpenCLGlobalFuncNames();
for (auto& func : funcmap)//Test if the functions the variation uses are actually included in those it says are required.
{
if (Find(str, func.first + "("))
{
if (!Contains(vec, func.first))
{
cout << "Variation " << var->Name() << " used global funcion " << func.first << ", but it's not found in its global func name vector: " << endl;
for (auto& v : vec)
cout << v << endl;
}
}
}
//Test whether the global functions the variations purports to need are actually used.
for (auto& v : vec)
{
bool found = false;
for (auto& v2 : vec)//Test if the functions the variation uses possibly use this function. It can be the case sometimes where a variation does not use it directly, but its global functions do.
{
if (v != v2)
{
auto it = funcmap.find(v2);
if (it != funcmap.end())
{
if (Find(it->second, v + "("))
{
found = true;
break;
}
}
}
}
if (!found && Find(str, v + "("))
{
found = true;
}
if (!found && Find(localfuncs, v + "("))
{
found = true;
}
if (!found)
{
cout << "Variation " << var->Name() << " purported to require the usage of global function " << v << ", but it's not found in its OpenCL function string:\n" /*<< kernel*/ << endl;
}
}
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
}
return success;
}
void PrintAllVars()
{
uint i = 0;
auto vlf(VariationList<float>::Instance());
while (auto var = vlf->GetVariation(i++))
cout << var->Name() << endl;
}
void TestXformsInOutPoints()
{
uint index = 0;
auto varList = VariationList<float>::Instance();
auto paletteList = PaletteList<float>::Instance();
QTIsaac<ISAAC_SIZE, ISAAC_INT> rand;
paletteList->Add("flam3-palettes.xml");
while (index < varList->RegSize())
{
vector<Xform<float>> xforms;
unique_ptr<Variation<float>> regVar(varList->GetVariationCopy(index, eVariationType::VARTYPE_REG));
string s = regVar->OpenCLString() + regVar->OpenCLFuncsString();
if (s.find("MwcNext") == string::npos)
{
unique_ptr<Variation<float>> preVar(varList->GetVariationCopy("pre_" + regVar->Name()));
unique_ptr<Variation<float>> postVar(varList->GetVariationCopy("post_" + regVar->Name()));
Xform<float> xform0(0.25f, rand.Frand01<float>(), rand.Frand11<float>(), 1, rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>());
Xform<float> xform1(0.25f, rand.Frand01<float>(), rand.Frand11<float>(), 1, rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>());
Xform<float> xform2(0.25f, rand.Frand01<float>(), rand.Frand11<float>(), 1, rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>());
Xform<float> xform3(0.25f, rand.Frand01<float>(), rand.Frand11<float>(), 1, rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>());
Xform<float> xform4(0.25f, rand.Frand01<float>(), rand.Frand11<float>(), 1, rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>());
Xform<float> xform5(0.25f, rand.Frand01<float>(), rand.Frand11<float>(), 1, rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>());
Xform<float> xform6(0.25f, rand.Frand01<float>(), rand.Frand11<float>(), 1, rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>());
Xform<float> xform7(0.25f, rand.Frand01<float>(), rand.Frand11<float>(), 1, rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>(), rand.Frand11<float>());
if (preVar.get() && postVar.get())
{
xform1.AddVariation(preVar->Copy());
xform2.AddVariation(regVar->Copy());
xform3.AddVariation(postVar->Copy());
xform4.AddVariation(preVar->Copy());
xform4.AddVariation(regVar->Copy());
xform5.AddVariation(preVar->Copy());
xform5.AddVariation(postVar->Copy());
xform6.AddVariation(regVar->Copy());
xform6.AddVariation(postVar->Copy());
xform7.AddVariation(preVar->Copy());
xform7.AddVariation(regVar->Copy());
xform7.AddVariation(postVar->Copy());
xforms.push_back(xform0);
xforms.push_back(xform1);
xforms.push_back(xform2);
xforms.push_back(xform3);
xforms.push_back(xform4);
xforms.push_back(xform5);
xforms.push_back(xform6);
xforms.push_back(xform7);
}
else
{
xform1.AddVariation(regVar->Copy());
xforms.push_back(xform0);
xforms.push_back(xform1);
}
for (size_t i = 0; i < xforms.size(); i++)
{
bool badVals = false;
Point<float> orig;
orig.m_X = rand.Frand11<float>();
orig.m_Y = rand.Frand11<float>();
orig.m_Z = rand.Frand11<float>();
orig.m_ColorX = rand.Frand01<float>();
orig.m_Opacity = rand.Frand01<float>();
Point<float> p1 = orig, p2 = orig, p3;
xforms[i].Apply(&p1, &p1, rand);
xforms[i].Apply(&p2, &p3, rand);
badVals |= (p1.m_X != p1.m_X);
badVals |= (p1.m_Y != p1.m_Y);
badVals |= (p1.m_Z != p1.m_Z);
badVals |= (p1.m_ColorX != p1.m_ColorX);
badVals |= (p1.m_Opacity != p1.m_Opacity);
badVals |= (p3.m_X != p3.m_X);
badVals |= (p3.m_Y != p3.m_Y);
badVals |= (p3.m_Z != p3.m_Z);
badVals |= (p3.m_ColorX != p3.m_ColorX);
badVals |= (p3.m_Opacity != p3.m_Opacity);
if (badVals)
cout << "Variation " << regVar->Name() << ": Bad value detected" << endl;
if (!badVals)
{
if (p1.m_X != p3.m_X)
cout << "Variation " << regVar->Name() << ": p1.m_X " << p1.m_X << " != p3.m_X " << p3.m_X << endl;
if (p1.m_Y != p3.m_Y)
cout << "Variation " << regVar->Name() << ": p1.m_Y " << p1.m_Y << " != p3.m_Y " << p3.m_Y << endl;
if (p1.m_Z != p3.m_Z)
cout << "Variation " << regVar->Name() << ": p1.m_Z " << p1.m_Z << " != p3.m_Z " << p3.m_Z << endl;
if (p1.m_ColorX != p3.m_ColorX)
cout << "Variation " << regVar->Name() << ": p1.m_ColorX " << p1.m_ColorX << " != p3.m_ColorX " << p3.m_ColorX << endl;
if (p1.m_Opacity != p3.m_Opacity)
cout << "Variation " << regVar->Name() << ": p1.m_Opacity " << p1.m_Opacity << " != p3.m_Opacity " << p3.m_Opacity << endl;
}
}
}
index++;
}
}
static int SortPairByTime(const pair<string, double>& a, pair<string, double>& b)
{
return a.second < b.second;
}
template <typename T>
void TestVarTime()
{
int i = 0, iters = 10;
Timing t;
VariationList<T> vlf;
IteratorHelper<T> helper;
QTIsaac<ISAAC_SIZE, ISAAC_INT> rand;
vector<pair<string, double>> times;
times.reserve(vlf->RegSize());
while (i < vlf->RegSize())
{
double sum = 0;
Xform<T> xform;
Variation<T>* var = vlf->GetVariationCopy(i, eVariationType::VARTYPE_REG);
xform.AddVariation(var);
for (int iter = 0; iter < iters; iter++)
{
Point<T> p;
xform.m_Affine.A(rand.Frand<T>(-5, 5));
xform.m_Affine.B(rand.Frand<T>(-5, 5));
xform.m_Affine.C(rand.Frand<T>(-5, 5));
xform.m_Affine.D(rand.Frand<T>(-5, 5));
xform.m_Affine.E(rand.Frand<T>(-5, 5));
xform.m_Affine.F(rand.Frand<T>(-5, 5));
p.m_X = rand.Frand<T>(-20, 20);
p.m_Y = rand.Frand<T>(-20, 20);
p.m_Z = rand.Frand<T>(-20, 20);
helper.In.x = helper.m_TransX = (xform.m_Affine.A() * p.m_X) + (xform.m_Affine.B() * p.m_Y) + xform.m_Affine.C();
helper.In.y = helper.m_TransY = (xform.m_Affine.D() * p.m_X) + (xform.m_Affine.E() * p.m_Y) + xform.m_Affine.F();
helper.In.z = helper.m_TransZ = p.m_Z;
helper.m_Color.x = p.m_ColorX = rand.Frand01<T>();
p.m_Opacity = rand.Frand01<T>();
helper.m_PrecalcSumSquares = SQR(helper.m_TransX) + SQR(helper.m_TransY);
helper.m_PrecalcSqrtSumSquares = std::sqrt(helper.m_PrecalcSumSquares);
helper.m_PrecalcSina = helper.m_TransX / helper.m_PrecalcSqrtSumSquares;
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
helper.m_PrecalcCosa = helper.m_TransY / helper.m_PrecalcSqrtSumSquares;
helper.m_PrecalcAtanxy = std::atan2(helper.m_TransX, helper.m_TransY);
helper.m_PrecalcAtanyx = std::atan2(helper.m_TransY, helper.m_TransX);
var->Random(rand);
t.Tic();
var->Func(helper, p, rand);
sum += t.Toc();
}
i++;
times.push_back(pair<string, double>(var->Name(), sum / iters));
}
std::sort(times.begin(), times.end(), &SortPairByTime);
//forr (auto& p : times) cout << p.first << "\t" << p.second << "" << endl;
}
void TestCasting()
{
vector<string> stringVec;
auto varList = VariationList<float>::Instance();
auto& vars = varList->AllVars();
stringVec.push_back("T(");
stringVec.push_back(".0f");
stringVec.push_back(".1f");
stringVec.push_back(".2f");
stringVec.push_back(".3f");
stringVec.push_back(".4f");
stringVec.push_back(".5f");
stringVec.push_back(".6f");
stringVec.push_back(".7f");
stringVec.push_back(".8f");
stringVec.push_back(".9f");
auto varVec = FindVarsWith<float>(vars, stringVec);
for (auto& it : varVec)
{
cout << "Variation " << it->Name() << " contained an improper float cast." << endl;
}
}
template <typename T>
void TestOperations()
{
vector<string> stringVec;
auto varList = VariationList<T>::Instance();
auto& vars = varList->AllVars();
//stringVec.push_back("%");
0.4.1.3 Beta 10/14/2014 --User Changes Size is no longer fixed to the window size. Size scaling is done differently in the final render dialog. This fixes several bugs. Remove Xml saving size from settings and options dialog, it no longer applies. Final render can be broken into strips. Set default save path to the desktop if none is found in the settings file. Set default output size to 1920x1080 if none is found in the settings file. --Bug Fixes Better memory size reporting in final render dialog. --Code Changes Migrate to C++11, Qt 5.3.1, and Visual Studio 2013. Change most instances of unsigned int to size_t, and int to intmax_t. Add m_OrigPixPerUnit and m_ScaleType to Ember for scaling purposes. Replace some sprintf_s() calls in XmlToEmber with ostringstream. Move more non-templated members into RendererBase. Add CopyVec() overload that takes a per element function pointer. Add vector Memset(). Replace '&' with '+' instead of "&amp;" in XmlToEmber for much faster parsing. Break strips rendering out into EmberCommon and call from EmberRender and Fractorium. Make AddAndWriteBuffer() just call WriteBuffer(). Make AddAndWriteImage() delete the existing image first before replacing it. Add SetOutputTexture() to RendererCL to support making new textures in response to resize events. Remove multiple return statements in RendererCL, and replace with a bool that tracks results. Add ToDouble(), MakeEnd(), ToString() and Exists() wrappers in Fractorium. Add Size() wrapper in EmberFile. Make QString function arguments const QString&, and string with const string&. Make ShowCritical() wrapper for invoking a message box from another thread. Add combo box to TwoButtonWidget and rename.
2014-10-14 11:53:15 -04:00
//varVec = FindVarsWith<T>(Vec);
//
//for (size_t i = 0; i < varVec.size(); i++)
//{
// cout << "Variation " << varVec[i]->Name() << " contained a modulo operation. Ensure its right hand operand is not zero." << endl;
//}
//
//stringVec.clear();
//ClearVec<Variation<T>>(varVec);
stringVec.push_back("MwcNext(mwc) %");
stringVec.push_back("MwcNext(mwc)%");
auto varVec = FindVarsWith<T>(vars, stringVec);
for (size_t i = 0; i < varVec.size(); i++)
{
cout << "Variation " << varVec[i]->Name() << " contained MwcNext(mwc) %. Use MwcNextRange() instead." << endl;
}
}
2019-06-11 22:19:20 -04:00
void TestArbitrary()
{
vector<string> stringVec, withoutVec;
2019-06-11 22:19:20 -04:00
auto varList = VariationList<float>::Instance();
auto& vars = varList->AllVars();
stringVec.push_back("switch");
//stringVec.push_back("if (!");
withoutVec.push_back("=");
withoutVec.push_back("<");
withoutVec.push_back(">");
//stringVec.push_back(" = vIn.x - ");
//stringVec.push_back(" = vIn.y - ");
//stringVec.push_back("sqrt(");
//stringVec.push_back("atan2(");
//stringVec.push_back("sin(");
//stringVec.push_back("cos(");
2019-06-11 22:19:20 -04:00
//stringVec.push_back("sincos(");
auto varVec = FindVarsWith<float>(vars, stringVec, true, true);
//auto varVec = FindVarsWithWithout<float>(vars, stringVec, withoutVec);
2019-06-11 22:19:20 -04:00
for (auto& it : varVec)
{
cout << "Variation " << it->Name() << " contained the desired w/wo strings." << endl;
2019-06-11 22:19:20 -04:00
}
}
template <typename T>
void TestVarsSimilar()
{
int i = 0, compIndex = 0, iters = 10;
Timing t;
VariationList<T> vlf;
IteratorHelper<T> helper;
QTIsaac<ISAAC_SIZE, ISAAC_INT> rand;
vector<pair<string, double>> diffs;
diffs.reserve(vlf->RegSize());
while (i < vlf->RegSize())
{
double diff = 0, highest = TMAX;
Xform<T> xform;
Variation<T>* var = vlf->GetVariationCopy(i, eVariationType::VARTYPE_REG);
pair<string, double> match("", TMAX);
compIndex = 0;
xform.AddVariation(var);
while (compIndex < vlf->RegSize())
{
if (compIndex == i)
{
compIndex++;
continue;
}
double sum = 0, xdiff = 0, ydiff = 0, zdiff = 0;
Xform<T> xformComp;
Variation<T>* varComp = vlf->GetVariationCopy(compIndex, eVariationType::VARTYPE_REG);
xformComp.AddVariation(varComp);
ParametricVariation<T>* parVar = dynamic_cast<ParametricVariation<T>*>(var);
ParametricVariation<T>* parVarComp = dynamic_cast<ParametricVariation<T>*>(varComp);
for (int iter = 0; iter < iters; iter++)
{
Point<T> p, pComp;
xform.m_Affine.A(rand.Frand<T>(-5, 5));
xform.m_Affine.B(rand.Frand<T>(-5, 5));
xform.m_Affine.C(rand.Frand<T>(-5, 5));
xform.m_Affine.D(rand.Frand<T>(-5, 5));
xform.m_Affine.E(rand.Frand<T>(-5, 5));
xform.m_Affine.F(rand.Frand<T>(-5, 5));
xformComp.m_Affine = xform.m_Affine;
p.m_X = rand.Frand<T>(-20, 20);
p.m_Y = rand.Frand<T>(-20, 20);
p.m_Z = rand.Frand<T>(-20, 20);
helper.In.x = helper.m_TransX = (xform.m_Affine.A() * p.m_X) + (xform.m_Affine.B() * p.m_Y) + xform.m_Affine.C();
helper.In.y = helper.m_TransY = (xform.m_Affine.D() * p.m_X) + (xform.m_Affine.E() * p.m_Y) + xform.m_Affine.F();
helper.In.z = helper.m_TransZ = p.m_Z;
helper.m_Color.x = p.m_ColorX = rand.Frand01<T>();
p.m_Opacity = rand.Frand01<T>();
pComp = p;
helper.m_PrecalcSumSquares = SQR(helper.m_TransX) + SQR(helper.m_TransY);
helper.m_PrecalcSqrtSumSquares = std::sqrt(helper.m_PrecalcSumSquares);
helper.m_PrecalcSina = helper.m_TransX / helper.m_PrecalcSqrtSumSquares;
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
helper.m_PrecalcCosa = helper.m_TransY / helper.m_PrecalcSqrtSumSquares;
helper.m_PrecalcAtanxy = std::atan2(helper.m_TransX, helper.m_TransY);
helper.m_PrecalcAtanyx = std::atan2(helper.m_TransY, helper.m_TransX);
if (parVar)
{
--User changes -Add support for multiple GPU devices. --These options are present in the command line and in Fractorium. -Change scheme of specifying devices from platform,device to just total device index. --Single number on the command line. --Change from combo boxes for device selection to a table of all devices in Fractorium. -Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Bug fixes -EmberAnimate, EmberRender, FractoriumSettings, FinalRenderDialog: Fix wrong order of arguments to Clamp() when assigning thread priority. -VariationsDC.h: Fix NVidia OpenCL compilation error in DCTriangleVariation. -FractoriumXformsColor.cpp: Checking for null pixmap pointer is not enough, must also check if the underlying buffer is null via call to QPixmap::isNull(). --Code changes -Ember.h: Add case for FLAME_MOTION_NONE and default in ApplyFlameMotion(). -EmberMotion.h: Call base constructor. -EmberPch.h: #pragma once only on Windows. -EmberToXml.h: --Handle different types of exceptions. --Add default cases to ToString(). -Isaac.h: Remove unused variable in constructor. -Point.h: Call base constructor in Color(). -Renderer.h/cpp: --Add bool to Alloc() to only allocate memory for the histogram. Needed for multi-GPU. --Make CoordMap() return a const ref, not a pointer. -SheepTools.h: --Use 64-bit types like the rest of the code already does. --Fix some comment misspellings. -Timing.h: Make BeginTime(), EndTime(), ElapsedTime() and Format() be const functions. -Utils.h: --Add new functions Equal() and Split(). --Handle more exception types in ReadFile(). --Get rid of most legacy blending of C and C++ argument parsing. -XmlToEmber.h: --Get rid of most legacy blending of C and C++ code from flam3. --Remove some unused variables. -EmberAnimate: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --If a render fails, exit since there is no point in continuing an animation with a missing frame. --Pass variables to threaded save better, which most likely fixes a very subtle bug that existed before. --Remove some unused variables. -EmberGenome, EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. -EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --Only print values when not rendering with OpenCL, since they're always 0 in that case. -EmberCLPch.h: --#pragma once only on Windows. --#include <atomic>. -IterOpenCLKernelCreator.h: Add new kernel for summing two histograms. This is needed for multi-GPU. -OpenCLWrapper.h: --Move all OpenCL info related code into its own class OpenCLInfo. --Add members to cache the values of global memory size and max allocation size. -RendererCL.h/cpp: --Redesign to accomodate multi-GPU. --Constructor now takes a vector of devices. --Remove DumpErrorReport() function, it's handled in the base. --ClearBuffer(), ReadPoints(), WritePoints(), ReadHist() and WriteHist() now optionally take a device index as a parameter. --MakeDmap() override and m_DmapCL member removed because it no longer applies since the histogram is always float since the last commit. --Add new function SumDeviceHist() to sum histograms from two devices by first copying to a temporary on the host, then a temporary on the device, then summing. --m_Calls member removed, as it's now per-device. --OpenCLWrapper removed. --m_Seeds member is now a vector of vector of seeds, to accomodate a separate and different array of seeds for each device. --Added member m_Devices, a vector of unique_ptr of RendererCLDevice. -EmberCommon.h --Added Devices() function to convert from a vector of device indices to a vector of platform,device indices. --Changed CreateRenderer() to accept a vector of devices to create a single RendererCL which will split work across multiple devices. --Added CreateRenderers() function to accept a vector of devices to create multiple RendererCL, each which will render on a single device. --Add more comments to some existing functions. -EmberCommonPch.h: #pragma once only on Windows. -EmberOptions.h: --Remove --platform option, it's just sequential device number now with the --device option. --Make --out be OPT_USE_RENDER instead of OPT_RENDER_ANIM since it's an error condition when animating. It makes no sense to write all frames to a single image. --Add Devices() function to parse comma separated --device option string and return a vector of device indices. --Make int and uint types be 64-bit, so intmax_t and size_t. --Make better use of macros. -JpegUtils.h: Make string parameters to WriteJpeg() and WritePng() be const ref. -All project files: Turn off buffer security check option in Visual Studio (/Gs-) -deployment.pri: Remove the line OTHER_FILES +=, it's pointless and was causing problems. -Ember.pro, EmberCL.pro: Add CONFIG += plugin, otherwise it wouldn't link. -EmberCL.pro: Add new files for multi-GPU support. -build_all.sh: use -j4 and QMAKE=${QMAKE:/usr/bin/qmake} -shared_settings.pri: -Add version string. -Remove old DESTDIR definitions. -Add the following lines or else nothing would build: CONFIG(release, debug|release) { CONFIG += warn_off DESTDIR = ../../../Bin/release } CONFIG(debug, debug|release) { DESTDIR = ../../../Bin/debug } QMAKE_POST_LINK += $$quote(cp --update ../../../Data/flam3-palettes.xml $${DESTDIR}$$escape_expand(\n\t)) LIBS += -L/usr/lib -lpthread -AboutDialog.ui: Another futile attempt to make it look correct on Linux. -FinalRenderDialog.h/cpp: --Add support for multi-GPU. --Change from combo boxes for device selection to a table of all devices. --Ensure device selection makes sense. --Remove "FinalRender" prefix of various function names, it's implied given the context. -FinalRenderEmberController.h/cpp: --Add support for multi-GPU. --Change m_FinishedImageCount to be atomic. --Move CancelRender() from the base to FinalRenderEmberController<T>. --Refactor RenderComplete() to omit any progress related functionality or image saving since it can be potentially ran in a thread. --Consolidate setting various renderer fields into SyncGuiToRenderer(). -Fractorium.cpp: Allow for resizing of the options dialog to show the entire device table. -FractoriumCommon.h: Add various functions to handle a table showing the available OpenCL devices on the system. -FractoriumEmberController.h/cpp: Remove m_FinalImageIndex, it's no longer needed. -FractoriumRender.cpp: Scale the interactive sub batch count and quality by the number of devices used. -FractoriumSettings.h/cpp: --Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Add multi-GPU support, remove old device,platform pair. -FractoriumToolbar.cpp: Disable OpenCL toolbar button if there are no devices present on the system. -FractoriumOptionsDialog.h/cpp: --Add support for multi-GPU. --Consolidate more assignments in DataToGui(). --Enable/disable CPU/OpenCL items in response to OpenCL checkbox event. -Misc: Convert almost everything to size_t for unsigned, intmax_t for signed.
2015-09-12 21:33:45 -04:00
for (size_t v = 0; v < parVar->ParamCount(); v++)
parVar->SetParamVal(v, (T)iter);
}
if (parVarComp)
{
--User changes -Add support for multiple GPU devices. --These options are present in the command line and in Fractorium. -Change scheme of specifying devices from platform,device to just total device index. --Single number on the command line. --Change from combo boxes for device selection to a table of all devices in Fractorium. -Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Bug fixes -EmberAnimate, EmberRender, FractoriumSettings, FinalRenderDialog: Fix wrong order of arguments to Clamp() when assigning thread priority. -VariationsDC.h: Fix NVidia OpenCL compilation error in DCTriangleVariation. -FractoriumXformsColor.cpp: Checking for null pixmap pointer is not enough, must also check if the underlying buffer is null via call to QPixmap::isNull(). --Code changes -Ember.h: Add case for FLAME_MOTION_NONE and default in ApplyFlameMotion(). -EmberMotion.h: Call base constructor. -EmberPch.h: #pragma once only on Windows. -EmberToXml.h: --Handle different types of exceptions. --Add default cases to ToString(). -Isaac.h: Remove unused variable in constructor. -Point.h: Call base constructor in Color(). -Renderer.h/cpp: --Add bool to Alloc() to only allocate memory for the histogram. Needed for multi-GPU. --Make CoordMap() return a const ref, not a pointer. -SheepTools.h: --Use 64-bit types like the rest of the code already does. --Fix some comment misspellings. -Timing.h: Make BeginTime(), EndTime(), ElapsedTime() and Format() be const functions. -Utils.h: --Add new functions Equal() and Split(). --Handle more exception types in ReadFile(). --Get rid of most legacy blending of C and C++ argument parsing. -XmlToEmber.h: --Get rid of most legacy blending of C and C++ code from flam3. --Remove some unused variables. -EmberAnimate: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --If a render fails, exit since there is no point in continuing an animation with a missing frame. --Pass variables to threaded save better, which most likely fixes a very subtle bug that existed before. --Remove some unused variables. -EmberGenome, EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. -EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --Only print values when not rendering with OpenCL, since they're always 0 in that case. -EmberCLPch.h: --#pragma once only on Windows. --#include <atomic>. -IterOpenCLKernelCreator.h: Add new kernel for summing two histograms. This is needed for multi-GPU. -OpenCLWrapper.h: --Move all OpenCL info related code into its own class OpenCLInfo. --Add members to cache the values of global memory size and max allocation size. -RendererCL.h/cpp: --Redesign to accomodate multi-GPU. --Constructor now takes a vector of devices. --Remove DumpErrorReport() function, it's handled in the base. --ClearBuffer(), ReadPoints(), WritePoints(), ReadHist() and WriteHist() now optionally take a device index as a parameter. --MakeDmap() override and m_DmapCL member removed because it no longer applies since the histogram is always float since the last commit. --Add new function SumDeviceHist() to sum histograms from two devices by first copying to a temporary on the host, then a temporary on the device, then summing. --m_Calls member removed, as it's now per-device. --OpenCLWrapper removed. --m_Seeds member is now a vector of vector of seeds, to accomodate a separate and different array of seeds for each device. --Added member m_Devices, a vector of unique_ptr of RendererCLDevice. -EmberCommon.h --Added Devices() function to convert from a vector of device indices to a vector of platform,device indices. --Changed CreateRenderer() to accept a vector of devices to create a single RendererCL which will split work across multiple devices. --Added CreateRenderers() function to accept a vector of devices to create multiple RendererCL, each which will render on a single device. --Add more comments to some existing functions. -EmberCommonPch.h: #pragma once only on Windows. -EmberOptions.h: --Remove --platform option, it's just sequential device number now with the --device option. --Make --out be OPT_USE_RENDER instead of OPT_RENDER_ANIM since it's an error condition when animating. It makes no sense to write all frames to a single image. --Add Devices() function to parse comma separated --device option string and return a vector of device indices. --Make int and uint types be 64-bit, so intmax_t and size_t. --Make better use of macros. -JpegUtils.h: Make string parameters to WriteJpeg() and WritePng() be const ref. -All project files: Turn off buffer security check option in Visual Studio (/Gs-) -deployment.pri: Remove the line OTHER_FILES +=, it's pointless and was causing problems. -Ember.pro, EmberCL.pro: Add CONFIG += plugin, otherwise it wouldn't link. -EmberCL.pro: Add new files for multi-GPU support. -build_all.sh: use -j4 and QMAKE=${QMAKE:/usr/bin/qmake} -shared_settings.pri: -Add version string. -Remove old DESTDIR definitions. -Add the following lines or else nothing would build: CONFIG(release, debug|release) { CONFIG += warn_off DESTDIR = ../../../Bin/release } CONFIG(debug, debug|release) { DESTDIR = ../../../Bin/debug } QMAKE_POST_LINK += $$quote(cp --update ../../../Data/flam3-palettes.xml $${DESTDIR}$$escape_expand(\n\t)) LIBS += -L/usr/lib -lpthread -AboutDialog.ui: Another futile attempt to make it look correct on Linux. -FinalRenderDialog.h/cpp: --Add support for multi-GPU. --Change from combo boxes for device selection to a table of all devices. --Ensure device selection makes sense. --Remove "FinalRender" prefix of various function names, it's implied given the context. -FinalRenderEmberController.h/cpp: --Add support for multi-GPU. --Change m_FinishedImageCount to be atomic. --Move CancelRender() from the base to FinalRenderEmberController<T>. --Refactor RenderComplete() to omit any progress related functionality or image saving since it can be potentially ran in a thread. --Consolidate setting various renderer fields into SyncGuiToRenderer(). -Fractorium.cpp: Allow for resizing of the options dialog to show the entire device table. -FractoriumCommon.h: Add various functions to handle a table showing the available OpenCL devices on the system. -FractoriumEmberController.h/cpp: Remove m_FinalImageIndex, it's no longer needed. -FractoriumRender.cpp: Scale the interactive sub batch count and quality by the number of devices used. -FractoriumSettings.h/cpp: --Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Add multi-GPU support, remove old device,platform pair. -FractoriumToolbar.cpp: Disable OpenCL toolbar button if there are no devices present on the system. -FractoriumOptionsDialog.h/cpp: --Add support for multi-GPU. --Consolidate more assignments in DataToGui(). --Enable/disable CPU/OpenCL items in response to OpenCL checkbox event. -Misc: Convert almost everything to size_t for unsigned, intmax_t for signed.
2015-09-12 21:33:45 -04:00
for (size_t v = 0; v < parVarComp->ParamCount(); v++)
parVarComp->SetParamVal(v, (T)iter);
}
//For debugging.
if (var->VariationId() == eVariationId::VAR_BWRAPS && varComp->VariationId() == eVariationId::VAR_ECLIPSE)
{
//cout << "Break." << endl;
}
helper.Out = v4T(0);
var->m_Weight = T(iter + 1);
var->Precalc();
var->Func(helper, p, rand);
v4T varOut = helper.Out;
helper.Out = v4T(0);
varComp->m_Weight = T(iter + 1);
varComp->Precalc();
varComp->Func(helper, pComp, rand);
v4T varCompOut = helper.Out;
xdiff += std::abs(varOut.x - varCompOut.x);
ydiff += std::abs(varOut.y - varCompOut.y);
zdiff += std::abs(varOut.z - varCompOut.z);
}
sum = (xdiff + ydiff + zdiff) / iters;
if (sum < highest)
{
match.first = varComp->Name();
match.second = highest = sum;
}
compIndex++;
}
if (match.second < 0.001)
cout << "The closest match to variation " << var->Name() << " is " << match.first << " with a total difference of " << match.second << endl;
i++;
//times.push_back(pair<string, double>(var->Name(), sum / iters));
}
//std::sort(times.begin(), times.end(), &SortPairByTime);
}
#ifdef TEST_CL
template <typename T>
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
bool TestAllVarsCLBuild(size_t platform, size_t device, bool printSuccess = true)
{
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
bool success = true;
vector<Ember<T>> embers;
QTIsaac<ISAAC_SIZE, ISAAC_INT> rand;
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
vector<pair<size_t, size_t>> devices{ std::make_pair(platform, device) };
RendererCL<T, float> renderer(devices);
const char* loc = __FUNCTION__;
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
if (!renderer.Ok())
{
cout << loc << "Creating RendererCL failed, tests will not be run." << endl;
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
return false;
}
MakeTestAllVarsRegPrePost(embers);
for (auto& it : embers)
{
renderer.SetEmber(it, eProcessAction::FULL_RENDER, true);
//if (platform != 0 &&
// ((it.GetXform(0)->GetVariationById(eVariationId::VAR_SYNTH) != nullptr) ||//Nvidia OpenCL driver crashes when building too many synths.
// (it.GetXform(0)->GetVariationById(eVariationId::VAR_PRE_SYNTH) != nullptr) ||
// (it.GetXform(0)->GetVariationById(eVariationId::VAR_POST_SYNTH) != nullptr)))
//{
// cout << "Skipping synth.\n";
// continue;
//}
2016-01-17 19:10:06 -05:00
if (renderer.BuildIterProgramForEmber())
{
if (printSuccess)
cout << loc << ": Build succeeded for ember " << it.m_Name << endl;
}
else
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
{
success = false;
cout << loc << ": OpenCL program build failed for ember " << it.m_Name << ":\n" << renderer.ErrorReportString() << endl;
break;
}
}
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
return success;
}
template <typename T>
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
void TestCpuGpuResults(size_t platform, size_t device)
{
bool breakOnBad = true;
int i = 0;//(int)eVariationId::VAR_TARGET;//Start at the one you want to test.
int iters = 10;
int skipped = 0;
T thresh = T(1e-3);
Timing t;
VariationList<T> vlf;
QTIsaac<ISAAC_SIZE, ISAAC_INT> rand;
vector<PointCL<T>> points;
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
vector<pair<size_t, size_t>> devices{ std::make_pair(platform, device) };
RendererCL<T, float> renderer(devices);
points.resize(renderer.TotalIterKernelCount());
while (i < vlf->RegSize())
{
bool bad = false;
double sum = 0;
Variation<T>* var = vlf->GetVariation(i, eVariationType::VARTYPE_REG);
string s = var->OpenCLString() + var->OpenCLFuncsString();
if (s.find("MwcNext") != string::npos)
{
i++;
skipped++;
continue;
}
cout << "Testing cpu-gpu equality for variation: " << var->Name() << " (" << (int)var->VariationId() << ")" << endl;
for (int iter = 0; iter < iters; iter++)
{
Point<T> p, p2;
Ember<T> ember;
Xform<T> xform;
Variation<T>* varCopy = var->Copy();
p.m_X = rand.Frand<T>(-5, 5);
p.m_Y = rand.Frand<T>(-5, 5);
p.m_Z = rand.Frand<T>(-5, 5);
p.m_ColorX = rand.Frand01<T>();
p.m_Opacity = rand.Frand01<T>();
varCopy->Random(rand);
xform.AddVariation(varCopy);
ember.AddXform(xform);
ember.CacheXforms();
renderer.SetEmber(ember, eProcessAction::FULL_RENDER, true);
renderer.AssignIterator();
if (!renderer.Alloc())
return;
points[0].m_X = p.m_X;
points[0].m_Y = p.m_Y;
points[0].m_Z = p.m_Z;
points[0].m_ColorX = p.m_ColorX;
xform.Apply(&p, &p2, rand);
renderer.WritePoints(points);
renderer.Iterate(1, 0, 1);
renderer.ReadPoints(points);
T xdiff = std::abs(p2.m_X - points[0].m_X);
T ydiff = std::abs(p2.m_Y - points[0].m_Y);
T zdiff = std::abs(p2.m_Z - points[0].m_Z);
if (xdiff > thresh || ydiff > thresh || zdiff > thresh)
{
bad = true;
cout << __FUNCTION__ << ": Variation cpu-gpu diff for iter " << iter << ": " << varCopy->Name() << " (" << (int)varCopy->VariationId() << ") xdiff: " << xdiff << endl;
cout << __FUNCTION__ << ": Variation cpu-gpu diff for iter " << iter << ": " << varCopy->Name() << " (" << (int)varCopy->VariationId() << ") ydiff: " << ydiff << endl;
cout << __FUNCTION__ << ": Variation cpu-gpu diff for iter " << iter << ": " << varCopy->Name() << " (" << (int)varCopy->VariationId() << ") zdiff: " << zdiff << endl;
cout << varCopy->ToString() << endl;
}
else
{
//cout << "Variation " << var->Name() << " had no difference between cpu and gpu for iter " << iter << endl;
}
}
if (breakOnBad && bad)
break;
i++;
bad = false;
}
cout << "Skipped " << skipped << endl;
}
template <typename T>
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
void TestGpuVectorRead(size_t platform, size_t device)
{
T minx = TMAX, miny = TMAX, minz = TMAX, mincolorx = TMAX;
T maxx = TLOW, maxy = TLOW, maxz = TLOW, maxcolorx = TLOW;
double sumx = 0, avgx = 0;
double sumy = 0, avgy = 0;
double sumz = 0, avgz = 0;
double sumcolorx = 0, avgcolorx = 0;
Timing t;
VariationList<T> vlf;
QTIsaac<ISAAC_SIZE, ISAAC_INT> rand;
vector<PointCL<T>> points;
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
vector<pair<size_t, size_t>> devices{ std::make_pair(platform, device) };
RendererCL<T, float> renderer(devices);
points.resize(renderer.TotalIterKernelCount());
Variation<T>* var = vlf->GetVariation(eVariationId::VAR_LINEAR);
Point<T> p, p2;
Ember<T> ember;
Xform<T> xform;
Variation<T>* varCopy = var->Copy();
p.m_X = rand.Frand<T>(-5, 5);
p.m_Y = rand.Frand<T>(-5, 5);
p.m_Z = rand.Frand<T>(-5, 5);
p.m_ColorX = rand.Frand01<T>();
p.m_Opacity = rand.Frand01<T>();
varCopy->Random(rand);
xform.AddVariation(varCopy);
ember.AddXform(xform);
ember.CacheXforms();
renderer.SetEmber(ember, eProcessAction::FULL_RENDER, true);
renderer.AssignIterator();
if (!renderer.Alloc())
return;
uint i, iters = renderer.IterCountPerKernel() * renderer.TotalIterKernelCount();//Make each thread in each block run at least once.
renderer.Iterate(iters, 0, 1);
renderer.ReadPoints(points);
cout << __FUNCTION__ << ": GPU point test value results:" << endl;
for (i = 0; i < points.size(); i++)
{
cout << "point[" << i << "].m_X = " << points[i].m_X << endl;
cout << "point[" << i << "].m_Y = " << points[i].m_Y << endl;
cout << "point[" << i << "].m_Z = " << points[i].m_Z << endl;
cout << "point[" << i << "].m_ColorX = " << points[i].m_ColorX << endl << endl;
minx = min<T>(points[i].m_X, minx);
miny = min<T>(points[i].m_Y, miny);
minz = min<T>(points[i].m_Z, minz);
mincolorx = min<T>(points[i].m_ColorX, mincolorx);
maxx = max<T>(points[i].m_X, maxx);
maxy = max<T>(points[i].m_Y, maxy);
maxz = max<T>(points[i].m_Z, maxz);
maxcolorx = max<T>(points[i].m_ColorX, maxcolorx);
sumx += points[i].m_X;
sumy += points[i].m_Y;
sumz += points[i].m_Z;
sumcolorx += points[i].m_ColorX;
}
avgx = sumx / i;
avgy = sumy / i;
avgz = sumz / i;
avgcolorx = sumcolorx / i;
cout << "avgx = " << avgx << endl;
cout << "avgy = " << avgy << endl;
cout << "avgz = " << avgz << endl;
cout << "avgcolorx = " << avgcolorx << endl;
cout << "minx = " << minx << endl;
cout << "miny = " << miny << endl;
cout << "minz = " << minz << endl;
cout << "mincolorx = " << mincolorx << endl << endl;
cout << "maxx = " << maxx << endl;
cout << "maxy = " << maxy << endl;
cout << "maxz = " << maxz << endl;
cout << "maxcolorx = " << maxcolorx << endl << endl << endl;
}
#endif
template <typename T>
void TestRandomAccess(size_t vsize, size_t ipp, bool cache)
{
size_t iters = vsize * ipp;
vector<v4T> vec;
QTIsaac<ISAAC_SIZE, ISAAC_INT> rand;
vec.resize(vsize);
v4T* vdata = vec.data();
if (cache)
{
for (size_t i = 0; i < iters; i++)
{
v4T v4(rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
int index = rand.Rand((ISAAC_INT)vsize);
v4T v42 = vdata[index];
v4.x = std::log(v4.x);
v4.y = std::sqrt(v4.y);
v4.z = std::sin(v4.z);
v4.w = std::cos(v4.w);
v4 += T(1.234);
v4 *= T(55.55);
v4 /= T(0.0045);
vdata[index] = v4 + v42;
}
}
else
{
for (size_t i = 0; i < iters; i++)
{
v4T v4(rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>(), rand.Frand11<T>());
int index = rand.Rand((ISAAC_INT)vsize);
v4.x = std::log(v4.x);
v4.y = std::sqrt(v4.y);
v4.z = std::sin(v4.z);
v4.w = std::cos(v4.w);
v4 += T(1.234);
v4 *= T(55.55);
v4 /= T(0.0045);
vdata[index] += v4;
}
}
}
template <typename T>
void TestCross(T x, T y, T weight)
{
T s = x * x - y * y;
T r = weight * std::sqrt(1 / (s * s + EPS));
T outX = x * r;
T outY = y * r;
cout << "First way, outX, outY == " << outX << ", " << outY << endl;
r = std::abs((x - y) * (x + y) + EPS);
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
if (r < 0)
r = -r;
r = weight / r;
outX = x * r;
outY = y * r;
cout << "Second way, outX, outY == " << outX << ", " << outY << endl;
}
double RandD(QTIsaac<ISAAC_SIZE, ISAAC_INT>& rand)
{
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
return ((((rand.Rand() ^ (rand.Rand() << 15)) & 0xfffffff) * 3.72529e-09) - 0.5);
}
//
//#define BEZ_POINT_LENGTH 4
//
//void BezierSolve(double t, glm::vec2* src, double* w, glm::vec2& solution)
//{
// double s, s2, s3, t2, t3, nom_x, nom_y, denom;
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
//
// s = 1 - t;
// s2 = s * s;
// s3 = s * s * s;
// t2 = t * t;
// t3 = t * t * t;
//
// nom_x = w[0] * s3 * src[0].x + w[1] * s2 * 3 * t * src[1].x + w[2] * s * 3 * t2 * src[2].x + w[3] * t3 * src[3].x;
//
// nom_y = w[0] * s3 * src[0].y + w[1] * s2 * 3 * t * src[1].y + w[2] * s * 3 * t2 * src[2].y + w[3] * t3 * src[3].y;
//
// denom = w[0] * s3 + w[1] * s2 * 3 * t + w[2] * s * 3 * t2 + w[3] * t3;
//
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
//
// if (isnan(nom_x) || isnan(nom_y) || isnan(denom) || denom == 0)
// return;
//
// solution.x = nom_x / denom;
// solution.y = nom_y / denom;
//}
//
//void BezierSetRect(glm::vec2* points, bool flip, glm::vec4& rect)
//{
// double f;
//
// for (int i = 0; i < BEZ_POINT_LENGTH; i++)
// {
// if (flip)
// f = 1 - points[i].y;
// else
// f = points[i].y;
//
// points[i].x = points[i].x * (rect.z - rect.x) + rect.x;
// points[i].y = f * (rect.w - rect.y) + rect.y;
// }
//}
//
//void BezierUnsetRect(glm::vec2* points, bool flip, glm::vec4& rect)
//{
// if ((rect.z - rect.x) == 0 || (rect.w - rect.y) == 0)
// return;
//
// for (int i = 0; i < BEZ_POINT_LENGTH; i++)
// {
// points[i].x = (points[i].x - rect.x) / (rect.z - rect.x);
// points[i].y = (points[i].y - rect.y) / (rect.w - rect.y);
//
// if (flip)
// points[i].y = 1 - points[i].y;
// }
//}
//
//struct BezierPoints
//{
// glm::vec2 points[4];
//};
//
//struct BezierWeights
//{
// double points[4];
//};
--User changes -Add support for multiple GPU devices. --These options are present in the command line and in Fractorium. -Change scheme of specifying devices from platform,device to just total device index. --Single number on the command line. --Change from combo boxes for device selection to a table of all devices in Fractorium. -Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Bug fixes -EmberAnimate, EmberRender, FractoriumSettings, FinalRenderDialog: Fix wrong order of arguments to Clamp() when assigning thread priority. -VariationsDC.h: Fix NVidia OpenCL compilation error in DCTriangleVariation. -FractoriumXformsColor.cpp: Checking for null pixmap pointer is not enough, must also check if the underlying buffer is null via call to QPixmap::isNull(). --Code changes -Ember.h: Add case for FLAME_MOTION_NONE and default in ApplyFlameMotion(). -EmberMotion.h: Call base constructor. -EmberPch.h: #pragma once only on Windows. -EmberToXml.h: --Handle different types of exceptions. --Add default cases to ToString(). -Isaac.h: Remove unused variable in constructor. -Point.h: Call base constructor in Color(). -Renderer.h/cpp: --Add bool to Alloc() to only allocate memory for the histogram. Needed for multi-GPU. --Make CoordMap() return a const ref, not a pointer. -SheepTools.h: --Use 64-bit types like the rest of the code already does. --Fix some comment misspellings. -Timing.h: Make BeginTime(), EndTime(), ElapsedTime() and Format() be const functions. -Utils.h: --Add new functions Equal() and Split(). --Handle more exception types in ReadFile(). --Get rid of most legacy blending of C and C++ argument parsing. -XmlToEmber.h: --Get rid of most legacy blending of C and C++ code from flam3. --Remove some unused variables. -EmberAnimate: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --If a render fails, exit since there is no point in continuing an animation with a missing frame. --Pass variables to threaded save better, which most likely fixes a very subtle bug that existed before. --Remove some unused variables. -EmberGenome, EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. -EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --Only print values when not rendering with OpenCL, since they're always 0 in that case. -EmberCLPch.h: --#pragma once only on Windows. --#include <atomic>. -IterOpenCLKernelCreator.h: Add new kernel for summing two histograms. This is needed for multi-GPU. -OpenCLWrapper.h: --Move all OpenCL info related code into its own class OpenCLInfo. --Add members to cache the values of global memory size and max allocation size. -RendererCL.h/cpp: --Redesign to accomodate multi-GPU. --Constructor now takes a vector of devices. --Remove DumpErrorReport() function, it's handled in the base. --ClearBuffer(), ReadPoints(), WritePoints(), ReadHist() and WriteHist() now optionally take a device index as a parameter. --MakeDmap() override and m_DmapCL member removed because it no longer applies since the histogram is always float since the last commit. --Add new function SumDeviceHist() to sum histograms from two devices by first copying to a temporary on the host, then a temporary on the device, then summing. --m_Calls member removed, as it's now per-device. --OpenCLWrapper removed. --m_Seeds member is now a vector of vector of seeds, to accomodate a separate and different array of seeds for each device. --Added member m_Devices, a vector of unique_ptr of RendererCLDevice. -EmberCommon.h --Added Devices() function to convert from a vector of device indices to a vector of platform,device indices. --Changed CreateRenderer() to accept a vector of devices to create a single RendererCL which will split work across multiple devices. --Added CreateRenderers() function to accept a vector of devices to create multiple RendererCL, each which will render on a single device. --Add more comments to some existing functions. -EmberCommonPch.h: #pragma once only on Windows. -EmberOptions.h: --Remove --platform option, it's just sequential device number now with the --device option. --Make --out be OPT_USE_RENDER instead of OPT_RENDER_ANIM since it's an error condition when animating. It makes no sense to write all frames to a single image. --Add Devices() function to parse comma separated --device option string and return a vector of device indices. --Make int and uint types be 64-bit, so intmax_t and size_t. --Make better use of macros. -JpegUtils.h: Make string parameters to WriteJpeg() and WritePng() be const ref. -All project files: Turn off buffer security check option in Visual Studio (/Gs-) -deployment.pri: Remove the line OTHER_FILES +=, it's pointless and was causing problems. -Ember.pro, EmberCL.pro: Add CONFIG += plugin, otherwise it wouldn't link. -EmberCL.pro: Add new files for multi-GPU support. -build_all.sh: use -j4 and QMAKE=${QMAKE:/usr/bin/qmake} -shared_settings.pri: -Add version string. -Remove old DESTDIR definitions. -Add the following lines or else nothing would build: CONFIG(release, debug|release) { CONFIG += warn_off DESTDIR = ../../../Bin/release } CONFIG(debug, debug|release) { DESTDIR = ../../../Bin/debug } QMAKE_POST_LINK += $$quote(cp --update ../../../Data/flam3-palettes.xml $${DESTDIR}$$escape_expand(\n\t)) LIBS += -L/usr/lib -lpthread -AboutDialog.ui: Another futile attempt to make it look correct on Linux. -FinalRenderDialog.h/cpp: --Add support for multi-GPU. --Change from combo boxes for device selection to a table of all devices. --Ensure device selection makes sense. --Remove "FinalRender" prefix of various function names, it's implied given the context. -FinalRenderEmberController.h/cpp: --Add support for multi-GPU. --Change m_FinishedImageCount to be atomic. --Move CancelRender() from the base to FinalRenderEmberController<T>. --Refactor RenderComplete() to omit any progress related functionality or image saving since it can be potentially ran in a thread. --Consolidate setting various renderer fields into SyncGuiToRenderer(). -Fractorium.cpp: Allow for resizing of the options dialog to show the entire device table. -FractoriumCommon.h: Add various functions to handle a table showing the available OpenCL devices on the system. -FractoriumEmberController.h/cpp: Remove m_FinalImageIndex, it's no longer needed. -FractoriumRender.cpp: Scale the interactive sub batch count and quality by the number of devices used. -FractoriumSettings.h/cpp: --Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Add multi-GPU support, remove old device,platform pair. -FractoriumToolbar.cpp: Disable OpenCL toolbar button if there are no devices present on the system. -FractoriumOptionsDialog.h/cpp: --Add support for multi-GPU. --Consolidate more assignments in DataToGui(). --Enable/disable CPU/OpenCL items in response to OpenCL checkbox event. -Misc: Convert almost everything to size_t for unsigned, intmax_t for signed.
2015-09-12 21:33:45 -04:00
void TestThreadedKernel()
{
OpenCLWrapper wrapper1, wrapper2;
if (wrapper1.Init(1, 0) && wrapper2.Init(2, 0))
{
string k = ConstantDefinesString(false) + "\n__kernel void Kern()\n"
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
"{\n"
" int gid = GLOBAL_ID_X + GLOBAL_ID_Y;\n"
"}\n"
"\n";
--User changes -Add support for multiple GPU devices. --These options are present in the command line and in Fractorium. -Change scheme of specifying devices from platform,device to just total device index. --Single number on the command line. --Change from combo boxes for device selection to a table of all devices in Fractorium. -Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Bug fixes -EmberAnimate, EmberRender, FractoriumSettings, FinalRenderDialog: Fix wrong order of arguments to Clamp() when assigning thread priority. -VariationsDC.h: Fix NVidia OpenCL compilation error in DCTriangleVariation. -FractoriumXformsColor.cpp: Checking for null pixmap pointer is not enough, must also check if the underlying buffer is null via call to QPixmap::isNull(). --Code changes -Ember.h: Add case for FLAME_MOTION_NONE and default in ApplyFlameMotion(). -EmberMotion.h: Call base constructor. -EmberPch.h: #pragma once only on Windows. -EmberToXml.h: --Handle different types of exceptions. --Add default cases to ToString(). -Isaac.h: Remove unused variable in constructor. -Point.h: Call base constructor in Color(). -Renderer.h/cpp: --Add bool to Alloc() to only allocate memory for the histogram. Needed for multi-GPU. --Make CoordMap() return a const ref, not a pointer. -SheepTools.h: --Use 64-bit types like the rest of the code already does. --Fix some comment misspellings. -Timing.h: Make BeginTime(), EndTime(), ElapsedTime() and Format() be const functions. -Utils.h: --Add new functions Equal() and Split(). --Handle more exception types in ReadFile(). --Get rid of most legacy blending of C and C++ argument parsing. -XmlToEmber.h: --Get rid of most legacy blending of C and C++ code from flam3. --Remove some unused variables. -EmberAnimate: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --If a render fails, exit since there is no point in continuing an animation with a missing frame. --Pass variables to threaded save better, which most likely fixes a very subtle bug that existed before. --Remove some unused variables. -EmberGenome, EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. -EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --Only print values when not rendering with OpenCL, since they're always 0 in that case. -EmberCLPch.h: --#pragma once only on Windows. --#include <atomic>. -IterOpenCLKernelCreator.h: Add new kernel for summing two histograms. This is needed for multi-GPU. -OpenCLWrapper.h: --Move all OpenCL info related code into its own class OpenCLInfo. --Add members to cache the values of global memory size and max allocation size. -RendererCL.h/cpp: --Redesign to accomodate multi-GPU. --Constructor now takes a vector of devices. --Remove DumpErrorReport() function, it's handled in the base. --ClearBuffer(), ReadPoints(), WritePoints(), ReadHist() and WriteHist() now optionally take a device index as a parameter. --MakeDmap() override and m_DmapCL member removed because it no longer applies since the histogram is always float since the last commit. --Add new function SumDeviceHist() to sum histograms from two devices by first copying to a temporary on the host, then a temporary on the device, then summing. --m_Calls member removed, as it's now per-device. --OpenCLWrapper removed. --m_Seeds member is now a vector of vector of seeds, to accomodate a separate and different array of seeds for each device. --Added member m_Devices, a vector of unique_ptr of RendererCLDevice. -EmberCommon.h --Added Devices() function to convert from a vector of device indices to a vector of platform,device indices. --Changed CreateRenderer() to accept a vector of devices to create a single RendererCL which will split work across multiple devices. --Added CreateRenderers() function to accept a vector of devices to create multiple RendererCL, each which will render on a single device. --Add more comments to some existing functions. -EmberCommonPch.h: #pragma once only on Windows. -EmberOptions.h: --Remove --platform option, it's just sequential device number now with the --device option. --Make --out be OPT_USE_RENDER instead of OPT_RENDER_ANIM since it's an error condition when animating. It makes no sense to write all frames to a single image. --Add Devices() function to parse comma separated --device option string and return a vector of device indices. --Make int and uint types be 64-bit, so intmax_t and size_t. --Make better use of macros. -JpegUtils.h: Make string parameters to WriteJpeg() and WritePng() be const ref. -All project files: Turn off buffer security check option in Visual Studio (/Gs-) -deployment.pri: Remove the line OTHER_FILES +=, it's pointless and was causing problems. -Ember.pro, EmberCL.pro: Add CONFIG += plugin, otherwise it wouldn't link. -EmberCL.pro: Add new files for multi-GPU support. -build_all.sh: use -j4 and QMAKE=${QMAKE:/usr/bin/qmake} -shared_settings.pri: -Add version string. -Remove old DESTDIR definitions. -Add the following lines or else nothing would build: CONFIG(release, debug|release) { CONFIG += warn_off DESTDIR = ../../../Bin/release } CONFIG(debug, debug|release) { DESTDIR = ../../../Bin/debug } QMAKE_POST_LINK += $$quote(cp --update ../../../Data/flam3-palettes.xml $${DESTDIR}$$escape_expand(\n\t)) LIBS += -L/usr/lib -lpthread -AboutDialog.ui: Another futile attempt to make it look correct on Linux. -FinalRenderDialog.h/cpp: --Add support for multi-GPU. --Change from combo boxes for device selection to a table of all devices. --Ensure device selection makes sense. --Remove "FinalRender" prefix of various function names, it's implied given the context. -FinalRenderEmberController.h/cpp: --Add support for multi-GPU. --Change m_FinishedImageCount to be atomic. --Move CancelRender() from the base to FinalRenderEmberController<T>. --Refactor RenderComplete() to omit any progress related functionality or image saving since it can be potentially ran in a thread. --Consolidate setting various renderer fields into SyncGuiToRenderer(). -Fractorium.cpp: Allow for resizing of the options dialog to show the entire device table. -FractoriumCommon.h: Add various functions to handle a table showing the available OpenCL devices on the system. -FractoriumEmberController.h/cpp: Remove m_FinalImageIndex, it's no longer needed. -FractoriumRender.cpp: Scale the interactive sub batch count and quality by the number of devices used. -FractoriumSettings.h/cpp: --Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Add multi-GPU support, remove old device,platform pair. -FractoriumToolbar.cpp: Disable OpenCL toolbar button if there are no devices present on the system. -FractoriumOptionsDialog.h/cpp: --Add support for multi-GPU. --Consolidate more assignments in DataToGui(). --Enable/disable CPU/OpenCL items in response to OpenCL checkbox event. -Misc: Convert almost everything to size_t for unsigned, intmax_t for signed.
2015-09-12 21:33:45 -04:00
if (wrapper1.AddProgram("prog1", k, "Kern", false) &&
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
wrapper2.AddProgram("prog1", k, "Kern", false))
--User changes -Add support for multiple GPU devices. --These options are present in the command line and in Fractorium. -Change scheme of specifying devices from platform,device to just total device index. --Single number on the command line. --Change from combo boxes for device selection to a table of all devices in Fractorium. -Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Bug fixes -EmberAnimate, EmberRender, FractoriumSettings, FinalRenderDialog: Fix wrong order of arguments to Clamp() when assigning thread priority. -VariationsDC.h: Fix NVidia OpenCL compilation error in DCTriangleVariation. -FractoriumXformsColor.cpp: Checking for null pixmap pointer is not enough, must also check if the underlying buffer is null via call to QPixmap::isNull(). --Code changes -Ember.h: Add case for FLAME_MOTION_NONE and default in ApplyFlameMotion(). -EmberMotion.h: Call base constructor. -EmberPch.h: #pragma once only on Windows. -EmberToXml.h: --Handle different types of exceptions. --Add default cases to ToString(). -Isaac.h: Remove unused variable in constructor. -Point.h: Call base constructor in Color(). -Renderer.h/cpp: --Add bool to Alloc() to only allocate memory for the histogram. Needed for multi-GPU. --Make CoordMap() return a const ref, not a pointer. -SheepTools.h: --Use 64-bit types like the rest of the code already does. --Fix some comment misspellings. -Timing.h: Make BeginTime(), EndTime(), ElapsedTime() and Format() be const functions. -Utils.h: --Add new functions Equal() and Split(). --Handle more exception types in ReadFile(). --Get rid of most legacy blending of C and C++ argument parsing. -XmlToEmber.h: --Get rid of most legacy blending of C and C++ code from flam3. --Remove some unused variables. -EmberAnimate: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --If a render fails, exit since there is no point in continuing an animation with a missing frame. --Pass variables to threaded save better, which most likely fixes a very subtle bug that existed before. --Remove some unused variables. -EmberGenome, EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. -EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --Only print values when not rendering with OpenCL, since they're always 0 in that case. -EmberCLPch.h: --#pragma once only on Windows. --#include <atomic>. -IterOpenCLKernelCreator.h: Add new kernel for summing two histograms. This is needed for multi-GPU. -OpenCLWrapper.h: --Move all OpenCL info related code into its own class OpenCLInfo. --Add members to cache the values of global memory size and max allocation size. -RendererCL.h/cpp: --Redesign to accomodate multi-GPU. --Constructor now takes a vector of devices. --Remove DumpErrorReport() function, it's handled in the base. --ClearBuffer(), ReadPoints(), WritePoints(), ReadHist() and WriteHist() now optionally take a device index as a parameter. --MakeDmap() override and m_DmapCL member removed because it no longer applies since the histogram is always float since the last commit. --Add new function SumDeviceHist() to sum histograms from two devices by first copying to a temporary on the host, then a temporary on the device, then summing. --m_Calls member removed, as it's now per-device. --OpenCLWrapper removed. --m_Seeds member is now a vector of vector of seeds, to accomodate a separate and different array of seeds for each device. --Added member m_Devices, a vector of unique_ptr of RendererCLDevice. -EmberCommon.h --Added Devices() function to convert from a vector of device indices to a vector of platform,device indices. --Changed CreateRenderer() to accept a vector of devices to create a single RendererCL which will split work across multiple devices. --Added CreateRenderers() function to accept a vector of devices to create multiple RendererCL, each which will render on a single device. --Add more comments to some existing functions. -EmberCommonPch.h: #pragma once only on Windows. -EmberOptions.h: --Remove --platform option, it's just sequential device number now with the --device option. --Make --out be OPT_USE_RENDER instead of OPT_RENDER_ANIM since it's an error condition when animating. It makes no sense to write all frames to a single image. --Add Devices() function to parse comma separated --device option string and return a vector of device indices. --Make int and uint types be 64-bit, so intmax_t and size_t. --Make better use of macros. -JpegUtils.h: Make string parameters to WriteJpeg() and WritePng() be const ref. -All project files: Turn off buffer security check option in Visual Studio (/Gs-) -deployment.pri: Remove the line OTHER_FILES +=, it's pointless and was causing problems. -Ember.pro, EmberCL.pro: Add CONFIG += plugin, otherwise it wouldn't link. -EmberCL.pro: Add new files for multi-GPU support. -build_all.sh: use -j4 and QMAKE=${QMAKE:/usr/bin/qmake} -shared_settings.pri: -Add version string. -Remove old DESTDIR definitions. -Add the following lines or else nothing would build: CONFIG(release, debug|release) { CONFIG += warn_off DESTDIR = ../../../Bin/release } CONFIG(debug, debug|release) { DESTDIR = ../../../Bin/debug } QMAKE_POST_LINK += $$quote(cp --update ../../../Data/flam3-palettes.xml $${DESTDIR}$$escape_expand(\n\t)) LIBS += -L/usr/lib -lpthread -AboutDialog.ui: Another futile attempt to make it look correct on Linux. -FinalRenderDialog.h/cpp: --Add support for multi-GPU. --Change from combo boxes for device selection to a table of all devices. --Ensure device selection makes sense. --Remove "FinalRender" prefix of various function names, it's implied given the context. -FinalRenderEmberController.h/cpp: --Add support for multi-GPU. --Change m_FinishedImageCount to be atomic. --Move CancelRender() from the base to FinalRenderEmberController<T>. --Refactor RenderComplete() to omit any progress related functionality or image saving since it can be potentially ran in a thread. --Consolidate setting various renderer fields into SyncGuiToRenderer(). -Fractorium.cpp: Allow for resizing of the options dialog to show the entire device table. -FractoriumCommon.h: Add various functions to handle a table showing the available OpenCL devices on the system. -FractoriumEmberController.h/cpp: Remove m_FinalImageIndex, it's no longer needed. -FractoriumRender.cpp: Scale the interactive sub batch count and quality by the number of devices used. -FractoriumSettings.h/cpp: --Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Add multi-GPU support, remove old device,platform pair. -FractoriumToolbar.cpp: Disable OpenCL toolbar button if there are no devices present on the system. -FractoriumOptionsDialog.h/cpp: --Add support for multi-GPU. --Consolidate more assignments in DataToGui(). --Enable/disable CPU/OpenCL items in response to OpenCL checkbox event. -Misc: Convert almost everything to size_t for unsigned, intmax_t for signed.
2015-09-12 21:33:45 -04:00
{
cout << "Builds ok, now run..." << endl;
std::thread th1([&]()
{
if (wrapper1.RunKernel(0, 256, 16, 1, 16, 16, 1))
{
cout << "Successful run inside thread 1..." << endl;
}
});
std::thread th2([&]()
{
if (wrapper2.RunKernel(0, 256, 16, 1, 16, 16, 1))
{
cout << "Successful run inside thread 2..." << endl;
}
});
Join(th1);
Join(th2);
--User changes -Add support for multiple GPU devices. --These options are present in the command line and in Fractorium. -Change scheme of specifying devices from platform,device to just total device index. --Single number on the command line. --Change from combo boxes for device selection to a table of all devices in Fractorium. -Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Bug fixes -EmberAnimate, EmberRender, FractoriumSettings, FinalRenderDialog: Fix wrong order of arguments to Clamp() when assigning thread priority. -VariationsDC.h: Fix NVidia OpenCL compilation error in DCTriangleVariation. -FractoriumXformsColor.cpp: Checking for null pixmap pointer is not enough, must also check if the underlying buffer is null via call to QPixmap::isNull(). --Code changes -Ember.h: Add case for FLAME_MOTION_NONE and default in ApplyFlameMotion(). -EmberMotion.h: Call base constructor. -EmberPch.h: #pragma once only on Windows. -EmberToXml.h: --Handle different types of exceptions. --Add default cases to ToString(). -Isaac.h: Remove unused variable in constructor. -Point.h: Call base constructor in Color(). -Renderer.h/cpp: --Add bool to Alloc() to only allocate memory for the histogram. Needed for multi-GPU. --Make CoordMap() return a const ref, not a pointer. -SheepTools.h: --Use 64-bit types like the rest of the code already does. --Fix some comment misspellings. -Timing.h: Make BeginTime(), EndTime(), ElapsedTime() and Format() be const functions. -Utils.h: --Add new functions Equal() and Split(). --Handle more exception types in ReadFile(). --Get rid of most legacy blending of C and C++ argument parsing. -XmlToEmber.h: --Get rid of most legacy blending of C and C++ code from flam3. --Remove some unused variables. -EmberAnimate: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --If a render fails, exit since there is no point in continuing an animation with a missing frame. --Pass variables to threaded save better, which most likely fixes a very subtle bug that existed before. --Remove some unused variables. -EmberGenome, EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. -EmberRender: --Support multi-GPU processing that alternates full frames between devices. --Use OpenCLInfo instead of OpenCLWrapper for --openclinfo option. --Remove bucketT template parameter, and hard code float in its place. --Only print values when not rendering with OpenCL, since they're always 0 in that case. -EmberCLPch.h: --#pragma once only on Windows. --#include <atomic>. -IterOpenCLKernelCreator.h: Add new kernel for summing two histograms. This is needed for multi-GPU. -OpenCLWrapper.h: --Move all OpenCL info related code into its own class OpenCLInfo. --Add members to cache the values of global memory size and max allocation size. -RendererCL.h/cpp: --Redesign to accomodate multi-GPU. --Constructor now takes a vector of devices. --Remove DumpErrorReport() function, it's handled in the base. --ClearBuffer(), ReadPoints(), WritePoints(), ReadHist() and WriteHist() now optionally take a device index as a parameter. --MakeDmap() override and m_DmapCL member removed because it no longer applies since the histogram is always float since the last commit. --Add new function SumDeviceHist() to sum histograms from two devices by first copying to a temporary on the host, then a temporary on the device, then summing. --m_Calls member removed, as it's now per-device. --OpenCLWrapper removed. --m_Seeds member is now a vector of vector of seeds, to accomodate a separate and different array of seeds for each device. --Added member m_Devices, a vector of unique_ptr of RendererCLDevice. -EmberCommon.h --Added Devices() function to convert from a vector of device indices to a vector of platform,device indices. --Changed CreateRenderer() to accept a vector of devices to create a single RendererCL which will split work across multiple devices. --Added CreateRenderers() function to accept a vector of devices to create multiple RendererCL, each which will render on a single device. --Add more comments to some existing functions. -EmberCommonPch.h: #pragma once only on Windows. -EmberOptions.h: --Remove --platform option, it's just sequential device number now with the --device option. --Make --out be OPT_USE_RENDER instead of OPT_RENDER_ANIM since it's an error condition when animating. It makes no sense to write all frames to a single image. --Add Devices() function to parse comma separated --device option string and return a vector of device indices. --Make int and uint types be 64-bit, so intmax_t and size_t. --Make better use of macros. -JpegUtils.h: Make string parameters to WriteJpeg() and WritePng() be const ref. -All project files: Turn off buffer security check option in Visual Studio (/Gs-) -deployment.pri: Remove the line OTHER_FILES +=, it's pointless and was causing problems. -Ember.pro, EmberCL.pro: Add CONFIG += plugin, otherwise it wouldn't link. -EmberCL.pro: Add new files for multi-GPU support. -build_all.sh: use -j4 and QMAKE=${QMAKE:/usr/bin/qmake} -shared_settings.pri: -Add version string. -Remove old DESTDIR definitions. -Add the following lines or else nothing would build: CONFIG(release, debug|release) { CONFIG += warn_off DESTDIR = ../../../Bin/release } CONFIG(debug, debug|release) { DESTDIR = ../../../Bin/debug } QMAKE_POST_LINK += $$quote(cp --update ../../../Data/flam3-palettes.xml $${DESTDIR}$$escape_expand(\n\t)) LIBS += -L/usr/lib -lpthread -AboutDialog.ui: Another futile attempt to make it look correct on Linux. -FinalRenderDialog.h/cpp: --Add support for multi-GPU. --Change from combo boxes for device selection to a table of all devices. --Ensure device selection makes sense. --Remove "FinalRender" prefix of various function names, it's implied given the context. -FinalRenderEmberController.h/cpp: --Add support for multi-GPU. --Change m_FinishedImageCount to be atomic. --Move CancelRender() from the base to FinalRenderEmberController<T>. --Refactor RenderComplete() to omit any progress related functionality or image saving since it can be potentially ran in a thread. --Consolidate setting various renderer fields into SyncGuiToRenderer(). -Fractorium.cpp: Allow for resizing of the options dialog to show the entire device table. -FractoriumCommon.h: Add various functions to handle a table showing the available OpenCL devices on the system. -FractoriumEmberController.h/cpp: Remove m_FinalImageIndex, it's no longer needed. -FractoriumRender.cpp: Scale the interactive sub batch count and quality by the number of devices used. -FractoriumSettings.h/cpp: --Temporal samples defaults to 100 instead of 1000 which was needless overkill. --Add multi-GPU support, remove old device,platform pair. -FractoriumToolbar.cpp: Disable OpenCL toolbar button if there are no devices present on the system. -FractoriumOptionsDialog.h/cpp: --Add support for multi-GPU. --Consolidate more assignments in DataToGui(). --Enable/disable CPU/OpenCL items in response to OpenCL checkbox event. -Misc: Convert almost everything to size_t for unsigned, intmax_t for signed.
2015-09-12 21:33:45 -04:00
cout << "Successful join of kernel thread..." << endl;
}
}
}
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
template <typename T>
void DistribTester()
{
size_t i;
size_t distribCount = 1;
size_t xformCount = 3;
vector<byte> m_XformDistributions;
size_t j = 0;
vector<T> weights { T(0.333333), T(1.0), T(0.25) };
double tempDensity = 0, currentDensityLimit = 0, densityPerElement;
if (m_XformDistributions.size() < CHOOSE_XFORM_GRAIN * distribCount)
m_XformDistributions.resize(CHOOSE_XFORM_GRAIN * distribCount);
if (m_XformDistributions.size() < CHOOSE_XFORM_GRAIN * distribCount)
return;
for (size_t distrib = 0; distrib < distribCount; distrib++)
{
double totalDensity = 0;
//First find the total densities of all xforms.
for (i = 0; i < xformCount; i++)
{
T d = weights[i];
totalDensity += d;
}
//Original returned false if all were 0, but it's allowed here
//which will just end up setting all elements to 0 which means
//only the first xform will get used.
//Calculate how much of a fraction of a the total density each element represents.
j = 0;
tempDensity = 0;
currentDensityLimit = 0;
densityPerElement = totalDensity / CHOOSE_XFORM_GRAIN;
//Assign xform indices in order to each element of m_XformDistributions.
//The number of elements assigned a given index is proportional to that xform's
//density relative to the sum of all densities.
for (i = 0; i < xformCount; i++)
{
T temp = weights[i];
currentDensityLimit += temp;
//Populate points corresponding to this xform's weight/density.
//Also check that j is within the bounds of the distribution array just to be safe in the case of a rounding error.
while (tempDensity < currentDensityLimit && j < CHOOSE_XFORM_GRAIN)
{
#ifdef _DEBUG
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
//Ensure distribution contains no out of bounds indices.
if (byte(i) >= xformCount)
throw "Out of bounds xform index in selection distribution.";
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
#endif
2016-12-05 22:04:33 -05:00
//cout << "offset = " << j << ", xform = " << i << ", running sum = " << tempDensity << "\n";
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
m_XformDistributions[(distrib * CHOOSE_XFORM_GRAIN) + j] = byte(i);
tempDensity += densityPerElement;
j++;
}
}
//Flam3 did this, which gives the same result.
//T t = xforms[0].m_Weight;
//
//if (distrib > 0)
// t *= xforms[distrib - 1].Xaos(0);
//
//T r = 0;
//
//for (i = 0; i < CHOOSE_XFORM_GRAIN; i++)
//{
// while (r >= t)
// {
// j++;
//
// if (distrib > 0)
// t += xforms[j].m_Weight * xforms[distrib - 1].Xaos(j);
// else
// t += xforms[j].m_Weight;
// }
//
// m_XformDistributions[(distrib * CHOOSE_XFORM_GRAIN) + i] = j;
// r += densityPerElement;
//}
}
}
template <typename T>
void TestAffine()
{
v2T x(1, 0), y(0, 1), t(1, 0);
Affine2D<T> af(x, y, t);
auto af2 = af;
cout << af.ToString() << "\n\n";
af.RotateTrans(90);
cout << af.ToString() << "\n\n";
af2.RotateTrans(-90);
cout << af2.ToString() << "\n\n";
}
template <typename T>
void TestRotate()
{
T angle = 45;
v3T x(1, 0, 0), y(0, 1, 0), xy(1, 1, 0);
auto xtrans = glm::translate(m4T(1), xy);
auto xrot = glm::rotate(xtrans, angle * DEG_2_RAD_T, v3T(0, 0, 1));
}
--User changes -Remove the option --intpalette to format the palette in the xml as ints. If they are not hex formatted, then they should always be float. This option was pointless. -Cleanup some options text for the command line programs. -Allow for dragging around flames in the library tab. This is useful for setting up the order of an animation. -Make the opening of large files in Fractorium much more efficient when not-appending. -Make the opening of large files in all EmberRender and EmberAnimate more efficient. -Better error reporting when opening files. --Bug fixes -Get rid of leftover artifacts that would appear on preview thumbnails when either switching SP/DP or re-rendering previews. -Filename extension was not being appended on Linux when saving as Xml, thus making it impossible to drag that file back in becase drop is filtered on extension. --Code changes -Move GCC compiler spec to C++14. Building with 5.3 now on linux. -Use inline member data initializers. -Make a #define for static for use in Utils.h to make things a little cleaner. -Make various functions able to take arbitrary collections as their parameters rather than just vectors. -Make library collection a list rather than vector. This alleviates the need to re-sync pointers whenever the collection changes. -Subclass QTreeWidget for the library tree. Two new files added for this. -Remove all usage of #ifdef ROW_ONLY_DE in DEOpenCLKernelCreator, it was never used. -Add move constructor and assignment operator to EmberFile. -Add the ability to use a pointer to outside memory in the renderer for the vector of Ember<T>. -Make a lot more functions const where they should be.
2016-04-03 21:55:12 -04:00
template<typename T>
class EmberContainerTester
{
public:
template <typename Alloc, template <typename, typename> class C>
static void TestEmberContainer(C<Ember<T>, Alloc>& cont)
{
Ember<T> e;
cont.push_back(e);
}
};
int _tmain(int argc, _TCHAR* argv[])
{
2014-09-10 01:41:26 -04:00
//int i;
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
bool b = true;
size_t times = 1'000'000'001;
QTIsaac<ISAAC_SIZE, ISAAC_INT> rand;
std::vector<unsigned int> vec(16, 0);
Timing t(4);
/*
for (size_t i = 1; i < times; i++)
{
auto res = rand.Rand() % i;
vec[res & 15]++;
}
t.Toc("rand mod");
for (auto& it : vec)
{
cout << it / (double)times << endl;
it = 0;
}
cout << "\n\n";
t.Tic();
for (size_t i = 1; i < times; i++)
{
//auto res = (uint)(((size_t)rand.Rand() * i) >> 32);
auto res = rand.Rand(i);
vec[res & 15]++;
}
t.Toc("rand mult shift");
for (auto& it : vec)
{
cout << it / (double)times << endl;
it = 0;
}
return 1;
vector<Ember<float>> fv;
vector<Ember<double>> dv;
list<Ember<float>> fl;
list<Ember<double>> dl;
int w = 1000, h = 1000;
string filename = ".\\testexr.exr";
vector<Rgba> pixels;
pixels.resize(w * h);
for (auto& pix : pixels)
{
pix.r = 1.0;
pix.b = 0.0;
pix.a = 1.0;
//pix.r = std::numeric_limits<float>::max();
}
writeRgba1(filename.c_str(), pixels.data(), w, h);
TestFuncs();
string line = "title=\"cj_aerie\" smooth=no", delim = " =\"";
auto vec = Split(line, delim, true);
for (auto& s : vec) cout << s << endl;
line = "index=0 color=2177354", delim = " =";
vec = Split(line, delim, true);
for (auto& s : vec) cout << s << endl;
EmberContainerTester<float>::TestEmberContainer(fv);
EmberContainerTester<double>::TestEmberContainer(dv);
EmberContainerTester<float>::TestEmberContainer(fl);
EmberContainerTester<double>::TestEmberContainer(dl);
CopyCont(fv, fl);
--User changes -Add a palette editor. -Add support for reading .ugr/.gradient/.gradients palette files. -Allow toggling on spinners whose minimum value is not zero. -Allow toggling display of image, affines and grid. -Add new variations: cylinder2, circlesplit, tile_log, truchet_fill, waves2_radial. --Bug fixes -cpow2 was wrong. -Palettes with rapid changes in color would produce slightly different outputs from Apo/Chaotica. This was due to a long standing bug from flam3. -Use exec() on Apple and show() on all other OSes for dialog boxes. -Trying to render a sequence with no frames would crash. -Selecting multiple xforms and rotating them would produce the wrong rotation. -Better handling when parsing flames using different encoding, such as unicode and UTF-8. -Switching between SP/DP didn't reselect the selected flame in the Library tab. --Code changes -Make all types concerning palettes be floats, including PaletteTableWidgetItem. -PaletteTableWidgetItem is no longer templated because all palettes are float. -Include the source colors for user created gradients. -Change parallel_for() calls to work with very old versions of TBB which are lingering on some systems. -Split conditional out of accumulation loop on the CPU for better performance. -Vectorize summing when doing density filter for better performance. -Make all usage of palettes be of type float, double is pointless. -Allow palettes to reside in multiple folders, while ensuring only one of each name is added. -Refactor some palette path searching code. -Make ReadFile() throw and catch an exception if the file operation fails. -A little extra safety in foci and foci3D with a call to Zeps(). -Cast to (real_t) in the OpenCL string for the w variation, which was having trouble compiling on Mac. -Fixing missing comma between paths in InitPaletteList(). -Move Xml and PaletteList classes into cpp to shorten build times when working on them. -Remove default param values for IterOpenCLKernelCreator<T>::SharedDataIndexDefines in cpp file. -Change more NULL to nullptr.
2017-02-26 03:02:21 -05:00
*/
--User changes -Remove the option --intpalette to format the palette in the xml as ints. If they are not hex formatted, then they should always be float. This option was pointless. -Cleanup some options text for the command line programs. -Allow for dragging around flames in the library tab. This is useful for setting up the order of an animation. -Make the opening of large files in Fractorium much more efficient when not-appending. -Make the opening of large files in all EmberRender and EmberAnimate more efficient. -Better error reporting when opening files. --Bug fixes -Get rid of leftover artifacts that would appear on preview thumbnails when either switching SP/DP or re-rendering previews. -Filename extension was not being appended on Linux when saving as Xml, thus making it impossible to drag that file back in becase drop is filtered on extension. --Code changes -Move GCC compiler spec to C++14. Building with 5.3 now on linux. -Use inline member data initializers. -Make a #define for static for use in Utils.h to make things a little cleaner. -Make various functions able to take arbitrary collections as their parameters rather than just vectors. -Make library collection a list rather than vector. This alleviates the need to re-sync pointers whenever the collection changes. -Subclass QTreeWidget for the library tree. Two new files added for this. -Remove all usage of #ifdef ROW_ONLY_DE in DEOpenCLKernelCreator, it was never used. -Add move constructor and assignment operator to EmberFile. -Add the ability to use a pointer to outside memory in the renderer for the vector of Ember<T>. -Make a lot more functions const where they should be.
2016-04-03 21:55:12 -04:00
//QTIsaac<ISAAC_SIZE, ISAAC_INT> rand(1, 2, 3);
//mt19937 meow(1729);
/* TestAffine<float>();
TestAffine<double>();*/
/* TestRotate<float>();
TestRotate<double>();
return 1;
*/
//MakeTestAllVarsRegPrePostComboFile("testallvarsout.flame");
//cout << (10.0 / 2.0 * 5.0) << endl;
//FindFmaCandidates();
//FindFmaImplemented();
//cout << 5 * 3 + 2 << endl;
//cout << 2 + 5 * 3 << endl;
/* return 0;
Features: --Add support for Exr files which use 32-bit floats for each RGBA channel. Seems to come out too washed out. --Allow for clearing an individual color curve. --Allow for saving multiple image types in EmberRender and EmberAnimate. All writes are threaded. --Remove --bpc command line argument. Add format png16 as a replacement. --Remove --enable_jpg_comments and --enable_png_comments command line arguments, and replace them with --enable_comments which applies to jpg, png and exr. --Add menu items to variations and affine spinners which allow for easy entry of specific numeric values like pi. --Make final render dialog be wider rather than so tall. Bug fixes: --Fix some OpenCL compile errors on Mac. --Remove ability to save bitmap files on all platforms but Windows. Code changes: --New dependency on OpenEXR. --Allow Curves class to interact with objects of a different template type. --Make m_Curves member of Ember always use float as template type. --Set the length of the curves array to always be 2^17 which should offer enough precision with new 32-bit float pixel types. --Set pixel types to always be 32-bit float. This results in a major reduction of code in the final accumulation part of Renderer.h/cpp. --Remove corresponding code from RendererCL and FinalAccumOpenCLKernelCreator. --Remove Transparency, NumChannels and BytesPerPixel setters from Renderer.h/cpp. --Add new global functions to format final image buffers and place all alpha calculation and scaling code in them. --Blending is no longer needed in OpenGLWidget because of the new pixel type. --Make new class, AffineDoubleSpinBox. --Attempt to make file save dialog code work the same on all OSes. --Remove some unused functions.
2017-07-22 16:43:35 -04:00
TestThreadedKernel();
Features: --Add support for Exr files which use 32-bit floats for each RGBA channel. Seems to come out too washed out. --Allow for clearing an individual color curve. --Allow for saving multiple image types in EmberRender and EmberAnimate. All writes are threaded. --Remove --bpc command line argument. Add format png16 as a replacement. --Remove --enable_jpg_comments and --enable_png_comments command line arguments, and replace them with --enable_comments which applies to jpg, png and exr. --Add menu items to variations and affine spinners which allow for easy entry of specific numeric values like pi. --Make final render dialog be wider rather than so tall. Bug fixes: --Fix some OpenCL compile errors on Mac. --Remove ability to save bitmap files on all platforms but Windows. Code changes: --New dependency on OpenEXR. --Allow Curves class to interact with objects of a different template type. --Make m_Curves member of Ember always use float as template type. --Set the length of the curves array to always be 2^17 which should offer enough precision with new 32-bit float pixel types. --Set pixel types to always be 32-bit float. This results in a major reduction of code in the final accumulation part of Renderer.h/cpp. --Remove corresponding code from RendererCL and FinalAccumOpenCLKernelCreator. --Remove Transparency, NumChannels and BytesPerPixel setters from Renderer.h/cpp. --Add new global functions to format final image buffers and place all alpha calculation and scaling code in them. --Blending is no longer needed in OpenGLWidget because of the new pixel type. --Make new class, AffineDoubleSpinBox. --Attempt to make file save dialog code work the same on all OSes. --Remove some unused functions.
2017-07-22 16:43:35 -04:00
auto palf = PaletteList<float>::Instance();
Palette<float>* pal = palf->GetRandomPalette();
Features: --Add support for Exr files which use 32-bit floats for each RGBA channel. Seems to come out too washed out. --Allow for clearing an individual color curve. --Allow for saving multiple image types in EmberRender and EmberAnimate. All writes are threaded. --Remove --bpc command line argument. Add format png16 as a replacement. --Remove --enable_jpg_comments and --enable_png_comments command line arguments, and replace them with --enable_comments which applies to jpg, png and exr. --Add menu items to variations and affine spinners which allow for easy entry of specific numeric values like pi. --Make final render dialog be wider rather than so tall. Bug fixes: --Fix some OpenCL compile errors on Mac. --Remove ability to save bitmap files on all platforms but Windows. Code changes: --New dependency on OpenEXR. --Allow Curves class to interact with objects of a different template type. --Make m_Curves member of Ember always use float as template type. --Set the length of the curves array to always be 2^17 which should offer enough precision with new 32-bit float pixel types. --Set pixel types to always be 32-bit float. This results in a major reduction of code in the final accumulation part of Renderer.h/cpp. --Remove corresponding code from RendererCL and FinalAccumOpenCLKernelCreator. --Remove Transparency, NumChannels and BytesPerPixel setters from Renderer.h/cpp. --Add new global functions to format final image buffers and place all alpha calculation and scaling code in them. --Blending is no longer needed in OpenGLWidget because of the new pixel type. --Make new class, AffineDoubleSpinBox. --Attempt to make file save dialog code work the same on all OSes. --Remove some unused functions.
2017-07-22 16:43:35 -04:00
cout << pal->Size() << endl;
Features: --Add support for Exr files which use 32-bit floats for each RGBA channel. Seems to come out too washed out. --Allow for clearing an individual color curve. --Allow for saving multiple image types in EmberRender and EmberAnimate. All writes are threaded. --Remove --bpc command line argument. Add format png16 as a replacement. --Remove --enable_jpg_comments and --enable_png_comments command line arguments, and replace them with --enable_comments which applies to jpg, png and exr. --Add menu items to variations and affine spinners which allow for easy entry of specific numeric values like pi. --Make final render dialog be wider rather than so tall. Bug fixes: --Fix some OpenCL compile errors on Mac. --Remove ability to save bitmap files on all platforms but Windows. Code changes: --New dependency on OpenEXR. --Allow Curves class to interact with objects of a different template type. --Make m_Curves member of Ember always use float as template type. --Set the length of the curves array to always be 2^17 which should offer enough precision with new 32-bit float pixel types. --Set pixel types to always be 32-bit float. This results in a major reduction of code in the final accumulation part of Renderer.h/cpp. --Remove corresponding code from RendererCL and FinalAccumOpenCLKernelCreator. --Remove Transparency, NumChannels and BytesPerPixel setters from Renderer.h/cpp. --Add new global functions to format final image buffers and place all alpha calculation and scaling code in them. --Blending is no longer needed in OpenGLWidget because of the new pixel type. --Make new class, AffineDoubleSpinBox. --Attempt to make file save dialog code work the same on all OSes. --Remove some unused functions.
2017-07-22 16:43:35 -04:00
double d = 1;
Features: --Add support for Exr files which use 32-bit floats for each RGBA channel. Seems to come out too washed out. --Allow for clearing an individual color curve. --Allow for saving multiple image types in EmberRender and EmberAnimate. All writes are threaded. --Remove --bpc command line argument. Add format png16 as a replacement. --Remove --enable_jpg_comments and --enable_png_comments command line arguments, and replace them with --enable_comments which applies to jpg, png and exr. --Add menu items to variations and affine spinners which allow for easy entry of specific numeric values like pi. --Make final render dialog be wider rather than so tall. Bug fixes: --Fix some OpenCL compile errors on Mac. --Remove ability to save bitmap files on all platforms but Windows. Code changes: --New dependency on OpenEXR. --Allow Curves class to interact with objects of a different template type. --Make m_Curves member of Ember always use float as template type. --Set the length of the curves array to always be 2^17 which should offer enough precision with new 32-bit float pixel types. --Set pixel types to always be 32-bit float. This results in a major reduction of code in the final accumulation part of Renderer.h/cpp. --Remove corresponding code from RendererCL and FinalAccumOpenCLKernelCreator. --Remove Transparency, NumChannels and BytesPerPixel setters from Renderer.h/cpp. --Add new global functions to format final image buffers and place all alpha calculation and scaling code in them. --Blending is no longer needed in OpenGLWidget because of the new pixel type. --Make new class, AffineDoubleSpinBox. --Attempt to make file save dialog code work the same on all OSes. --Remove some unused functions.
2017-07-22 16:43:35 -04:00
for (int i = 0; i < 10; i++)
{
cout << "log10(" << d << ") = " << std::max<uint>(1u, uint(std::log10(d)) + 1u) << endl;
d *= 10;
}
Features: --Add support for Exr files which use 32-bit floats for each RGBA channel. Seems to come out too washed out. --Allow for clearing an individual color curve. --Allow for saving multiple image types in EmberRender and EmberAnimate. All writes are threaded. --Remove --bpc command line argument. Add format png16 as a replacement. --Remove --enable_jpg_comments and --enable_png_comments command line arguments, and replace them with --enable_comments which applies to jpg, png and exr. --Add menu items to variations and affine spinners which allow for easy entry of specific numeric values like pi. --Make final render dialog be wider rather than so tall. Bug fixes: --Fix some OpenCL compile errors on Mac. --Remove ability to save bitmap files on all platforms but Windows. Code changes: --New dependency on OpenEXR. --Allow Curves class to interact with objects of a different template type. --Make m_Curves member of Ember always use float as template type. --Set the length of the curves array to always be 2^17 which should offer enough precision with new 32-bit float pixel types. --Set pixel types to always be 32-bit float. This results in a major reduction of code in the final accumulation part of Renderer.h/cpp. --Remove corresponding code from RendererCL and FinalAccumOpenCLKernelCreator. --Remove Transparency, NumChannels and BytesPerPixel setters from Renderer.h/cpp. --Add new global functions to format final image buffers and place all alpha calculation and scaling code in them. --Blending is no longer needed in OpenGLWidget because of the new pixel type. --Make new class, AffineDoubleSpinBox. --Attempt to make file save dialog code work the same on all OSes. --Remove some unused functions.
2017-07-22 16:43:35 -04:00
return 0;*/
/*
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
uint i, iters = (uint)10e7;
size_t total = 0;
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
t.Tic();
for (i = 0; i < iters; i++)
{
total += rand.RandByte();
total += rand.Rand();
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
}
t.Toc("Isaac sum");
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
cout << "Isaac total = " << total << " for " << i << " iters." << endl;
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
total = 0;
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
t.Tic();
for (i = 0; i < iters; i++)
{
total += meow();
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
}
t.Toc("Mt sum");
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
cout << "Mt total = " << total << " for " << i << " iters." << endl;
*/
//glm::vec2 solution, src[4];
//double bezT = 1, w[4];
//BezierPoints curvePoints[4];
//BezierWeights curveWeights[4];
//
//BezierSolve(bezT, src, w, solution);
//cout << pow(-1, 5.1) << endl;
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
/* for (i = 0; i < 2500000000; i++)
{
double d = std::abs(RandD(rand));
if (d >= 0.5)
cout << d << endl;
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
}
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
return 0;*/
//cout << "sizeof(Ember<float>): " << sizeof(Ember<float>) << endl;
//cout << "sizeof(Ember<double>): " << sizeof(Ember<double>) << endl;
//
//cout << "sizeof(Renderer<float>): " << sizeof(Renderer<float, float>) << endl;
//cout << "sizeof(Renderer<double>): " << sizeof(Renderer<double, double>) << endl;
//
//cout << "sizeof(RendererCL<float>): " << sizeof(RendererCL<float>) << endl;
//cout << "sizeof(RendererCL<double>): " << sizeof(RendererCL<double>) << endl;
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
/* unique_ptr<LinearVariation<float>> linV(new LinearVariation<float>());
unique_ptr<PreLinearVariation<float>> preLinV(new PreLinearVariation<float>());
unique_ptr<PostLinearVariation<float>> postLinV(new PostLinearVariation<float>());
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
cout << linV->BaseName() << endl;
cout << preLinV->BaseName() << endl;
cout << postLinV->BaseName() << endl;*/
//float num = 1;
//float denom = 4294967296.0f;
//float frac = num / denom;
//
//cout << "num, denom, frac = " << num << ", " << denom << ", " << frac << endl;
//TestGpuVectorRead<double>();
//TestGpuVectorRead<float>();
//return 0;
//unique_ptr<PreFarblurVariation<float>> preFarblurV(new PreFarblurVariation<float>());
//size_t vsize = 1024 * 1024;
//
//t.Tic();
//TestRandomAccess<float>(vsize, 10, true);
//t.Toc("TestRandomAccess<float>(true)");
//
//t.Tic();
//TestRandomAccess<float>(vsize, 10, false);
//t.Toc("TestRandomAccess<float>(false)");
//
//t.Tic();
//TestRandomAccess<double>(vsize, 10, true);
//t.Toc("TestRandomAccess<double>(true)");
//
//t.Tic();
//TestRandomAccess<double>(vsize, 10, false);
//t.Toc("TestRandomAccess<double>(false)");
//TestCross<double>(rand.Frand<double>(-5, 5), rand.Frand<double>(-5, 5), rand.Frand<double>(-5, 5));
//TestCross<double>(rand.Frand<double>(-5, 5), rand.Frand<double>(-5, 5), rand.Frand<double>(-5, 5));
//TestCross<double>(rand.Frand<double>(-5, 5), rand.Frand<double>(-5, 5), rand.Frand<double>(-5, 5));
//TestCross<double>(rand.Frand<double>(-5, 5), rand.Frand<double>(-5, 5), rand.Frand<double>(-5, 5));
//TestCross<double>(rand.Frand<double>(-5, 5), rand.Frand<double>(-5, 5), rand.Frand<double>(-5, 5));
//std::complex<double> cd, cd2;
//cd2 = sin(cd);
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
t.Tic();
TestCasting();
t.Toc("TestCasting()");
t.Tic();
auto vlf(VariationList<float>::Instance());
t.Toc("Creating VariationList<float>");
cout << "There are " << vlf->Size() << " variations present." << endl;
#ifdef DO_DOUBLE
t.Tic();
auto vld(VariationList<double>::Instance());
t.Toc("Creating VariationList<double>");
#endif
t.Tic();
TestVarCounts();
t.Toc("TestVarCounts()");
t.Tic();
TestVarUnique<float>();
t.Toc("TestVarUnique<float>()");
#ifdef DO_DOUBLE
t.Tic();
TestVarUnique<double>();
t.Toc("TestVarUnique<double>()");
#endif
t.Tic();
TestVarCopy<float, float>();
t.Toc("TestVarCopy<float, float>()");
#ifdef DO_DOUBLE
t.Tic();
TestVarCopy<double, double>();
t.Toc("TestVarCopy<double, double>()");
t.Tic();
TestVarCopy<float, double>();
t.Toc("TestVarCopy<float, double>()");
t.Tic();
TestVarCopy<double, float>();
t.Toc("TestVarCopy<double, float>()");
#endif
t.Tic();
TestVarRegPrePost();
t.Toc("TestVarRegPrePost()");
t.Tic();
TestParVars();
t.Toc("TestParVars()");
t.Tic();
TestVarPrePostNames();
t.Toc("TestVarPrePostNames()");
t.Tic();
TestVarPrecalcUsedCL();
t.Toc("TestVarPrecalcUsedCL()");
t.Tic();
TestVarAssignTypes();
t.Toc("TestVarAssignTypes()");
t.Tic();
TestVarAssignVals();
t.Toc("TestVarAssignVals()");
t.Tic();
TestZepsFloor();
t.Toc("TestZepsFloor()");
t.Tic();
TestConstants();
t.Toc("TestConstants()");
t.Tic();
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
TestGlobalFuncs();
t.Toc("TestGlobalFuncs()");
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
/* t.Tic();
TestXformsInOutPoints();
t.Toc("TestXformsInOutPoints()");
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
--User changes -Add new variations: crackle, dc_perlin. -Make default palette interp mode be linear instead of step. -Make summary tab the selected one in the Info tab. -Allow for highlight power of up to 10. It was previously limited to 2. --Bug fixes -Direct color calculations were wrong. -Flattening was not applied to final xform. -Fix "pure virtual function call" error on shutdown. --Code changes -Allow for array precalc params in variations by adding a size member to the ParamWithName class. -In IterOpenCLKernelCreator, memcpy precalc params instead of a direct assign since they can now be of variable length. -Add new file VarFuncs to consolidate some functions that are common to multiple variations. This also contains texture data for crackle and dc_perlin. -Place OpenCL versions of these functions in the FunctionMapper class in the EmberCL project. -Add new Singleton class that uses CRTP, is thread safe, and deletes after the last reference goes away. This fixes the usual "delete after main()" problem with singletons that use the static local function variable pattern. -Began saving files with AStyle autoformatter turned on. This will eventually touch all files as they are worked on. -Add missing backslash to CUDA include and library paths for builds on Nvidia systems. -Add missing gl.h include for Windows. -Remove glew include paths from Fractorium, it's not used. -Remove any Nvidia specific #defines and build targets, they are no longer needed with OpenCL 1.2. -Fix bad paths on linux build. -General cleanup.
2015-12-31 16:41:59 -05:00
t.Tic();
TestVarTime<float>();
t.Toc("TestVarTime()");
*/
t.Tic();
TestOperations<float>();
t.Toc("TestOperations()");
2019-06-11 22:19:20 -04:00
t.Tic();
TestArbitrary();
t.Toc("TestArbitrary()");
//t.Tic();
//TestVarsSimilar<float>();
//t.Toc("TestVarsSimilar()");
#ifdef TEST_CL
//t.Tic();
//TestCpuGpuResults<float>();
//t.Toc("TestCpuGpuResults<float>()");
//t.Tic();
//b = TestAllVarsCLBuild<float>(0, 0, true);
//t.Toc("TestAllVarsCLBuild<float>()");
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
if (b)
{
2016-01-17 19:10:06 -05:00
#ifdef DO_NVIDIA
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
t.Tic();
b = TestAllVarsCLBuild<float>(1, 0, true);
t.Toc("TestAllVarsCLBuild<float>()");
2016-01-17 19:10:06 -05:00
#endif
--User changes -Add new variations: bubbleT3D, crob, hexaplay3D, hexcrop, hexes, hexnix3D, loonie2, loonie3, nBlur, octapol and synth. -Allow for pre/post versions of dc_bubble, dc_cylinder and dc_linear whereas before they were omitted. -When saving a file with multiple embers in it, detect if time values are all the same and if so, start them at zero and increment by 1 for each ember. -Allow for numerous quality increases to be coalesced into one. It will pick up at the end of the current render. -Show selection highlight on variations tree in response to mouse hover. This makes it easier to see for which variation or param the current mouse wheel action will apply. -Make default temporal samples be 100, whereas before it was 1000 which was overkill. -Require the shift key to be held with delete for deleting an ember to prevent it from triggering when the user enters delete in the edit box. -This wasn't otherwise fixable without writing a lot more code. --Bug fixes -EmberGenome was crashing when generating a sequence from a source file with more than 2 embers in it. -EmberGenome was improperly handling the first frame of a merge after the last frame of the loop. -These bugs were due to a previous commit. Revert parts of that commit. -Prevent a zoom value of less than 0 when reading from xml. -Slight optimization of the crescents, and mask variations, if the compiler wasn't doing it already. -Unique file naming was broken because it was looking for _# and the default names ended with -#. -Disallow renaming of an ember in the library tree to an empty string. -Severe bug that prevented some variations from being read correctly from params generated outside this program. -Severe OpenCL randomization bug. The first x coordinates of the first points in the first kernel call of the first ember of a render since the OpenCL renderer object was created were not random and were mostly -1. -Severe bug when populating xform selection distributions that could sometimes cause a crash due to roundoff error. Fix by using double. -Limit the max number of variations in a random ember to MAX_CL_VARS, which is 8. This ensures they'll look the same on CPU and GPU. -Prevent user from saving stylesheet to default.qss, it's a special reserved filename. --Code changes -Generalize using the running sum output point inside of a variation for all cases: pre, reg and post. -Allow for array variables in variations where the address of each element is stored in m_Params. -Qualify all math functions with std:: -No longer use our own Clamp() in OpenCL, instead use the standard clamp(). -Redesign how functions are used in the variations OpenCL code. -Add tests to EmberTester to verify some of the new functionality. -Place more const and override qualifiers on functions where appropriate. -Add a global rand with a lock to be used very sparingly. -Use a map instead of a vector for bad param names in Xml parsing. -Prefix affine interpolation mode defines with "AFFINE_" to make their purpose more clear. -Allow for variations that change state during iteration by sending a separate copy of the ember to each rendering thread. -Implement this same functionality with a local struct in OpenCL. It's members are the total of all variables that need to change state within an ember. -Add Contains() function to Utils.h. -EmberRender: print names of kernels being printed with --dump_kernel option. -Clean up EmberTester to handle some of the recent changes. -Fix various casts. -Replace % 2 with & 1, even though the compiler was likely doing this already. -Add new file Variations06.h to accommodate new variations. -General cleanup.
2015-11-22 17:15:07 -05:00
}
#ifdef DO_DOUBLE
//t.Tic();
//TestCpuGpuResults<double>();
//t.Toc("TestCpuGpuResults<double>()");
if (b)
{
//t.Tic();
//b = TestAllVarsCLBuild<double>(0, 0, true);
//t.Toc("TestAllVarsCLBuild<double>()");
if (b)
{
#ifdef DO_NVIDIA
t.Tic();
TestAllVarsCLBuild<double>(1, 0, true);
t.Toc("TestAllVarsCLBuild<double>()");
#endif
}
}
#endif
#endif
//PrintAllVars();
//_CrtDumpMemoryLeaks();
return 0;
}