archive/blender-archive
Archived
1
1
Fork 0
Code Pull Requests Packages Activity
This repository has been archived on 2023-10-09. You can view files and clone it, but cannot push or open issues or pull requests.
Files
46f1c1b15d80535b8fb914f6a227384866bb2caf
blender-archive/source/blender/gpu/intern/gpu_extensions.c

432 lines
12 KiB
C
Raw Normal View History

doxygen: prevent GPL license block from being parsed as doxygen comment.
2011-02-23 10:52:22 +00:00
/*
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
remove header text: "The Blender Foundation also sells licenses for use in proprietary software under the Blender Licens" also remove NaN references from files that have been added since blender went opensource.
2011-11-29 10:54:47 +00:00
* of the License, or (at your option) any later version.
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software Foundation,
correct fsf address
2010-02-12 13:34:04 +00:00
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
*
* The Original Code is Copyright (C) 2005 Blender Foundation.
* All rights reserved.
*/
doxygen: add newline after \file While \file doesn't need an argument, it can't have another doxy command after it.
2019-02-18 08:08:12 +11:00
/** \file
* \ingroup gpu
Docs: doxygen file descriptions for BLF, GPU and WM
2014-01-19 23:14:24 +11:00
*
* Wrap OpenGL features such as textures, shaders and GLSL
* with checks for drivers and GPU support.
doxygen: blender/gpu tagged.
2011-02-27 20:25:53 +00:00
*/
split BKE_utildefines.h, now it only has blender specific defines like GS() MAKE_ID, FILE_MAXDIR, moved the generic defines to BLI_utildefines.h. no functional changes.
2011-01-07 18:36:47 +00:00
#include "BLI_utildefines.h"
static functions for getting power of 2 values were being copied about too much, add to the BLI_math api. - is_power_of_2_i - power_of_2_min_i - power_of_2_max_i
2011-12-16 09:25:07 +00:00
#include "BLI_math_base.h"
Potential fix for T43987, ambient occlusion different between offscreen and on screen rendering. Aaaaah, the beauty of driver implementations of OpenGL! Turns out the problem here is that drivers calculate df/dy differently in some cases (probably because OpenGL counts y reverse to how the window system does, so drivers can get confused). Fixed this for the ATI case based on info we have so far, there's also the Intel case which will be handled separately (missing info on Intel's renderer string etc). Unfortunately we can't really fix this for the general case so we'll have to haldle cases as they come in our tracker and by adding silly string comparisons in our GPU initialization module <sigh>.
2015-03-30 14:14:32 +02:00
#include "BLI_math_vector.h"
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
header cleanup, include BLI before BKE, also use bool for ntreeShaderExecTree
2013-01-24 21:57:13 +00:00
#include "BKE_global.h"
GPU: Fix Issue with recursive downsample and Intel HDXXX This is caused by a driver bug that prevent us from rendering to (or even binding) a texture mip level that is below GL_TEXTURE_MAX_LEVEL of the target texture. This is fine in most drivers (and legal AFAIK) but not on thoses Intels HDXXX + Windows. As a fix we just put GL_TEXTURE_MAX_LEVEL lower (which is illegal because it is undefined behaviour), but in practice it works ok and does not trigger any warnings or errors. This commit fixes most of the problems encountered on these GPUs (T56668).
2018-10-22 12:58:11 +02:00
#include "MEM_guardedalloc.h"
header cleanup, include BLI before BKE, also use bool for ntreeShaderExecTree
2013-01-24 21:57:13 +00:00
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
#include "GPU_extensions.h"
GPU: Fix Issue with recursive downsample and Intel HDXXX This is caused by a driver bug that prevent us from rendering to (or even binding) a texture mip level that is below GL_TEXTURE_MAX_LEVEL of the target texture. This is fine in most drivers (and legal AFAIK) but not on thoses Intels HDXXX + Windows. As a fix we just put GL_TEXTURE_MAX_LEVEL lower (which is illegal because it is undefined behaviour), but in practice it works ok and does not trigger any warnings or errors. This commit fixes most of the problems encountered on these GPUs (T56668).
2018-10-22 12:58:11 +02:00
#include "GPU_framebuffer.h"
OpenGL: split off framebuffer, shader and texture code into separate files.
2015-12-06 21:20:19 +01:00
#include "GPU_glew.h"
#include "GPU_texture.h"
Ghost Context Refactor https://developer.blender.org/D643 Separates graphics context creation from window code in Ghost so that they can vary separately.
2014-10-07 15:46:19 -05:00
Debug GPU functionality from soc-viewport_fx by Jason Wilkins patch number D706 with changes: - WITH_GPU_DEBUG just creates a debug context (and enables the debug messaging system functions) but leaves the checks we had intact. Old patch added the debug functionality only if we had the flag on to save some performance. Rationale here is that we might not want to recompile blender just to get the extra information, and having users start blender with a -d flag to get the extra information is also useful for bug reports. Those checks already existed and most expensive ones are hidden behind a debug mode check so performance should not be that bad. - Did some cleanup of existing functionality: When things go wrong blender side, just print the error, don't check for GL errors first. - Did not port changes needed for GLES to regular glew.h - Got rid of duplicate or very similar new functionality. Generally, code is more moving things around/cleanup and should work exactly as before apart from the debug context, so it's safe to add even now. It also provides a nice substitute function for glu error descriptions
2015-02-16 21:19:12 +01:00
#include "intern/gpu_private.h"
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
code cleanup: dont include BLI_winstuff.h on non windows systems, also cleanup some defines/includes for windows.
2012-04-15 07:54:07 +00:00
#ifdef WIN32
# include "BLI_winstuff.h"
#endif
Ensure gpu_extensions.c compiles with MSVC -> include BLI_winstuff.h
2011-03-07 00:49:21 +00:00
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
/* Extensions support */
OpenGL: bump gpu_extensions to GL 2.1, prepare for 3.x Several changes. Tested working on Windows 10 GL 4.3 and MacOS 10.11 GL 2.1. - document extensions used in this file - some simple ARB/EXT suffix deletion - stop checking for pre-2.1 features — they’re available! - convert old ARB shader API to the one adopted in GL 2.0 - remove checks for old (pre-R600) ATI cards - choose GLSL version at runtime, between 1.2 and 1.5 - prefer GLSL 1.5 for geometry shaders, fall back to EXT_geometry_shader4 if needed Differential Revision: https://developer.blender.org/D1632
2015-11-22 13:22:47 -05:00
/* -- extension: version of GL that absorbs it
OpenGL: transition to ARB FBOs, remove runtime checks ARB_framebuffer_object replaces several related EXT extensions. The ARB version pulls GL 3 FBO features into GL 2.1, useful for Mac platform. Its functions and enums have no ARB suffix so transition to modern GL will be seamless! Extension is checked at startup, so is guaranteed to be true at runtime. Part of T49012
2016-08-04 21:11:45 -04:00
* EXT_gpu_shader4: 3.0
OpenGL: bump gpu_extensions to GL 2.1, prepare for 3.x Several changes. Tested working on Windows 10 GL 4.3 and MacOS 10.11 GL 2.1. - document extensions used in this file - some simple ARB/EXT suffix deletion - stop checking for pre-2.1 features — they’re available! - convert old ARB shader API to the one adopted in GL 2.0 - remove checks for old (pre-R600) ATI cards - choose GLSL version at runtime, between 1.2 and 1.5 - prefer GLSL 1.5 for geometry shaders, fall back to EXT_geometry_shader4 if needed Differential Revision: https://developer.blender.org/D1632
2015-11-22 13:22:47 -05:00
* ARB_framebuffer object: 3.0
* EXT_framebuffer_multisample_blit_scaled: ???
* ARB_draw_instanced: 3.1
* ARB_texture_multisample: 3.2
* ARB_texture_query_lod: 4.0
style cleanup: comment blocks
2012-03-09 18:28:30 +00:00
*/
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
Added a -noglsl option to disable GLSL from the command line.
2008-10-12 18:32:26 +00:00
static struct GPUGlobal {
Add function to query maximum texture size. Also, make texture upload functions aware of this limit.
2013-04-12 17:56:07 +00:00
GLint maxtexsize;
Fix T55888: Eevee: crash when shadow cube size is > 512px Note that this was only reported to happen on AMD GPU + windows.
2018-07-27 13:56:07 +02:00
GLint maxtexlayers;
World textures displaying for viewport in BI. This patch supports "Image or Movie" and "Environment map" types of world texture for the viewport. It supports: - "View", "AngMap" and "Equirectangular" types of mapping. - Different types of texture blending (according to BI world render). - Same color blending as when it lacked textures (but render via glsl). {F207734} {F207735} Example: {F275180} Original author: @valentin_b4w Regards, Alexander (Blend4Web Team). Reviewers: sergey, valentin_b4w, brecht, merwin Reviewed By: merwin Subscribers: campbellbarton, merwin, blueprintrandom, youle, a.romanov, yurikovelenov, AlexKowel, Evgeny_Rodygin Projects: #rendering, #opengl_gfx, #bf_blender:_next Differential Revision: https://developer.blender.org/D1414
2016-01-27 12:06:57 +03:00
GLint maxcubemapsize;
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
GLint maxtextures;
GPUShader: Manually validate sampler count This happens on NVidia GPUs, using more textures than the maximum allowed by the gl will NOT trigger a linking issue (maybe because of bindless texture implementation?). So in this case we manually count the number of samplers per shader stage and compare it against the GL limit. We discard the shader if the sampler count is too high. This shows the user something is wrong with the shader.
2018-09-18 14:22:42 +02:00
GLint maxtexturesfrag;
GLint maxtexturesgeom;
GLint maxtexturesvert;
Clay-Engine (merge clay-engine) Initial work by Clément Foucault with contributions from Dalai Felinto (mainly per-collection engine settings logic, and depsgraph iterator placeholder). This makes Blender require OpenGL 3.3. Which means Intel graphic card and OSX will break. Disable CLAY_ENGINE in CMake in those cases. This is a prototype render engine intended to help the design of real render engines. This is mainly an engine with enphasis in matcap and ambient occlusion. Implemented Features -------------------- * Clay Render Engine, following the new API, to be used as reference for future engines * A more complete Matcap customization with more options * Per-Collection render engine settings * New Ground Truth AO - not enabled Missing Features ---------------- * Finish object edit mode - Fix shaders to use new matrix - Fix artifacts when edge does off screen - Fix depth issue - Selection sillhouette - Mesh wires - Use mesh normals (for higher quality matcap) - Non-Mesh objects drawing - Widget drawing - Performance issues * Finish mesh edit mode - Derived-Mesh-less edit mode API (mesh_rende.c) * General edit mode - Per-collection edit mode settings * General engines - Per-collection engine settings (they are their, but they still need to be flushed by depsgraph, and used by the drawing code)
2017-02-07 11:20:15 +01:00
GLint maxubosize;
GLint maxubobinds;
Fix #21894: backface selection wasn't working correct with < 24 bits colors, e.g. thousands of colors on OS X, due to use of uninitialized value. Problem tracked down and patch provided by Shane Ambler, thanks!
2010-07-04 12:24:19 +00:00
int colordepth;
Support for multi-sample off-screen buffers Replaces much slower manual accumulation buffer which simply did multiple renders. Needs OpenGL3.2, otherwise multi-sample is disabled.
2015-10-14 10:33:30 +11:00
int samples_color_texture_max;
Cleanup: use eGPU prefix for GPU enum types
2019-01-23 14:15:43 +11:00
eGPUDeviceType device;
eGPUOSType os;
eGPUDriverType driver;
GPU: Add safety check for max line width On some platform does not support line width > 1.0 and can even throw and error. Better check an at least display something rather than no lines at all.
2018-11-02 14:58:49 +01:00
float line_width_range[2];
Cleanup: comment blocks
2016-07-15 02:36:21 +10:00
/* workaround for different calculation of dfdy factors on GPUs. Some GPUs/drivers
* calculate dfdy in shader differently when drawing to an offscreen buffer. First
* number is factor on screen and second is off-screen */
float dfdyfactors[2];
OpenGL: fix max texture anisotropy check Query max supported on init, use that to clamp user-set values.
2016-01-07 11:21:08 -05:00
float max_anisotropy;
GPU: Fix Issue with recursive downsample and Intel HDXXX This is caused by a driver bug that prevent us from rendering to (or even binding) a texture mip level that is below GL_TEXTURE_MAX_LEVEL of the target texture. This is fine in most drivers (and legal AFAIK) but not on thoses Intels HDXXX + Windows. As a fix we just put GL_TEXTURE_MAX_LEVEL lower (which is illegal because it is undefined behaviour), but in practice it works ok and does not trigger any warnings or errors. This commit fixes most of the problems encountered on these GPUs (T56668).
2018-10-22 12:58:11 +02:00
/* Some Intel drivers have issues with using mips as framebuffer targets if
* GL_TEXTURE_MAX_LEVEL is higher than the target mip.
* We need a workaround in this cases. */
bool mip_render_workaround;
GPU: Add workarounds for buggy glBlitFramebuffer function on macOS + Radeon When calling glBlitFramebuffer on most (if not all) mac that have a GPU from the Radeon Pro series, the data is not properly copied and only a subset of the pixels are correctly copied. This only happens when blitting the depth buffer if the depth buffer is GL_DEPTH24_STENCIL8. Changing the depth buffer format to GPU_DEPTH32F_STENCIL8 fixes the issue but only works if blitting the depth componnent. The stencil componnent still provoke issues when being copied.
2018-10-25 21:57:40 +02:00
/* There is an issue with the glBlitFramebuffer on MacOS with radeon pro graphics.
* Blitting depth with GL_DEPTH24_STENCIL8 is buggy so the workaround is to use
* GPU_DEPTH32F_STENCIL8. Then Blitting depth will work but blitting stencil will
* still be broken. */
bool depth_blitting_workaround;
Workbench: Fix workbench broken on some config due to usuned fb slot This seems to be a driver bug. Only windows + Radeon HD 7500M seems to be affected. Fix can be extended to more config if necessary.
2018-12-04 20:29:30 +01:00
/* Crappy driver don't know how to map framebuffer slot to output vars...
* We need to have no "holes" in the output buffer slots. */
bool unused_fb_slot_workaround;
Merging phase 1 of the BGE Harmony branch: * Shadow color now usable in the BGE * Simplified the shadow panel while "Blender Game" renderer is active * Added variance shadow maps for the BGE * Buffered shadows on sun lamps in the BGE (orthographic) * Light textures in the BGE
2012-05-01 02:50:17 +00:00
} GG = {1, 0};
Added a -noglsl option to disable GLSL from the command line.
2008-10-12 18:32:26 +00:00
GPU: Fix Issue with recursive downsample and Intel HDXXX This is caused by a driver bug that prevent us from rendering to (or even binding) a texture mip level that is below GL_TEXTURE_MAX_LEVEL of the target texture. This is fine in most drivers (and legal AFAIK) but not on thoses Intels HDXXX + Windows. As a fix we just put GL_TEXTURE_MAX_LEVEL lower (which is illegal because it is undefined behaviour), but in practice it works ok and does not trigger any warnings or errors. This commit fixes most of the problems encountered on these GPUs (T56668).
2018-10-22 12:58:11 +02:00
static void gpu_detect_mip_render_workaround(void)
{
int cube_size = 2;
float *source_pix = MEM_callocN(sizeof(float) * 4 * 6 * cube_size * cube_size, __func__);
float clear_color[4] = {1.0f, 0.5f, 0.0f, 0.0f};
GPUTexture *tex = GPU_texture_create_cube(cube_size, GPU_RGBA16F, source_pix, NULL);
MEM_freeN(source_pix);
GPU_texture_bind(tex, 0);
GPU_texture_generate_mipmap(tex);
glTexParameteri(GPU_texture_target(tex), GL_TEXTURE_BASE_LEVEL, 0);
glTexParameteri(GPU_texture_target(tex), GL_TEXTURE_MAX_LEVEL, 0);
GPU_texture_unbind(tex);
GPUFrameBuffer *fb = GPU_framebuffer_create();
GPU_framebuffer_texture_attach(fb, tex, 0, 1);
GPU_framebuffer_bind(fb);
GPU_framebuffer_clear_color(fb, clear_color);
GPU_framebuffer_restore();
GPU_framebuffer_free(fb);
float *data = GPU_texture_read(tex, GPU_DATA_FLOAT, 1);
GG.mip_render_workaround = !equals_v4v4(clear_color, data);
MEM_freeN(data);
GPU_texture_free(tex);
}
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
/* GPU Types */
Cleanup: use eGPU prefix for GPU enum types
2019-01-23 14:15:43 +11:00
bool GPU_type_matches(eGPUDeviceType device, eGPUOSType os, eGPUDriverType driver)
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
{
return (GG.device & device) && (GG.os & os) && (GG.driver & driver);
}
/* GPU Extensions */
code cleanup: warnings + style
2013-04-13 00:43:49 +00:00
int GPU_max_texture_size(void)
Add function to query maximum texture size. Also, make texture upload functions aware of this limit.
2013-04-12 17:56:07 +00:00
{
return GG.maxtexsize;
}
Fix T55888: Eevee: crash when shadow cube size is > 512px Note that this was only reported to happen on AMD GPU + windows.
2018-07-27 13:56:07 +02:00
int GPU_max_texture_layers(void)
{
return GG.maxtexlayers;
}
OpenGL: split off framebuffer, shader and texture code into separate files.
2015-12-06 21:20:19 +01:00
int GPU_max_textures(void)
{
return GG.maxtextures;
}
GPUShader: Manually validate sampler count This happens on NVidia GPUs, using more textures than the maximum allowed by the gl will NOT trigger a linking issue (maybe because of bindless texture implementation?). So in this case we manually count the number of samplers per shader stage and compare it against the GL limit. We discard the shader if the sampler count is too high. This shows the user something is wrong with the shader.
2018-09-18 14:22:42 +02:00
int GPU_max_textures_frag(void)
{
return GG.maxtexturesfrag;
}
int GPU_max_textures_geom(void)
{
return GG.maxtexturesgeom;
}
int GPU_max_textures_vert(void)
{
return GG.maxtexturesvert;
}
OpenGL: fix max texture anisotropy check Query max supported on init, use that to clamp user-set values.
2016-01-07 11:21:08 -05:00
float GPU_max_texture_anisotropy(void)
{
return GG.max_anisotropy;
}
OpenGL: split off framebuffer, shader and texture code into separate files.
2015-12-06 21:20:19 +01:00
int GPU_max_color_texture_samples(void)
{
return GG.samples_color_texture_max;
}
World textures displaying for viewport in BI. This patch supports "Image or Movie" and "Environment map" types of world texture for the viewport. It supports: - "View", "AngMap" and "Equirectangular" types of mapping. - Different types of texture blending (according to BI world render). - Same color blending as when it lacked textures (but render via glsl). {F207734} {F207735} Example: {F275180} Original author: @valentin_b4w Regards, Alexander (Blend4Web Team). Reviewers: sergey, valentin_b4w, brecht, merwin Reviewed By: merwin Subscribers: campbellbarton, merwin, blueprintrandom, youle, a.romanov, yurikovelenov, AlexKowel, Evgeny_Rodygin Projects: #rendering, #opengl_gfx, #bf_blender:_next Differential Revision: https://developer.blender.org/D1414
2016-01-27 12:06:57 +03:00
int GPU_max_cube_map_size(void)
{
return GG.maxcubemapsize;
}
Clay-Engine (merge clay-engine) Initial work by Clément Foucault with contributions from Dalai Felinto (mainly per-collection engine settings logic, and depsgraph iterator placeholder). This makes Blender require OpenGL 3.3. Which means Intel graphic card and OSX will break. Disable CLAY_ENGINE in CMake in those cases. This is a prototype render engine intended to help the design of real render engines. This is mainly an engine with enphasis in matcap and ambient occlusion. Implemented Features -------------------- * Clay Render Engine, following the new API, to be used as reference for future engines * A more complete Matcap customization with more options * Per-Collection render engine settings * New Ground Truth AO - not enabled Missing Features ---------------- * Finish object edit mode - Fix shaders to use new matrix - Fix artifacts when edge does off screen - Fix depth issue - Selection sillhouette - Mesh wires - Use mesh normals (for higher quality matcap) - Non-Mesh objects drawing - Widget drawing - Performance issues * Finish mesh edit mode - Derived-Mesh-less edit mode API (mesh_rende.c) * General edit mode - Per-collection edit mode settings * General engines - Per-collection engine settings (they are their, but they still need to be flushed by depsgraph, and used by the drawing code)
2017-02-07 11:20:15 +01:00
int GPU_max_ubo_binds(void)
{
return GG.maxubobinds;
}
int GPU_max_ubo_size(void)
{
return GG.maxubosize;
}
GPU: Add safety check for max line width On some platform does not support line width > 1.0 and can even throw and error. Better check an at least display something rather than no lines at all.
2018-11-02 14:58:49 +01:00
float GPU_max_line_width(void)
{
return GG.line_width_range[1];
}
Potential fix for T43987, ambient occlusion different between offscreen and on screen rendering. Aaaaah, the beauty of driver implementations of OpenGL! Turns out the problem here is that drivers calculate df/dy differently in some cases (probably because OpenGL counts y reverse to how the window system does, so drivers can get confused). Fixed this for the ATI case based on info we have so far, there's also the Intel case which will be handled separately (missing info on Intel's renderer string etc). Unfortunately we can't really fix this for the general case so we'll have to haldle cases as they come in our tracker and by adding silly string comparisons in our GPU initialization module <sigh>.
2015-03-30 14:14:32 +02:00
void GPU_get_dfdy_factors(float fac[2])
{
copy_v2_v2(fac, GG.dfdyfactors);
}
GPU: Fix Issue with recursive downsample and Intel HDXXX This is caused by a driver bug that prevent us from rendering to (or even binding) a texture mip level that is below GL_TEXTURE_MAX_LEVEL of the target texture. This is fine in most drivers (and legal AFAIK) but not on thoses Intels HDXXX + Windows. As a fix we just put GL_TEXTURE_MAX_LEVEL lower (which is illegal because it is undefined behaviour), but in practice it works ok and does not trigger any warnings or errors. This commit fixes most of the problems encountered on these GPUs (T56668).
2018-10-22 12:58:11 +02:00
bool GPU_mip_render_workaround(void)
{
return GG.mip_render_workaround;
}
GPU: Add workarounds for buggy glBlitFramebuffer function on macOS + Radeon When calling glBlitFramebuffer on most (if not all) mac that have a GPU from the Radeon Pro series, the data is not properly copied and only a subset of the pixels are correctly copied. This only happens when blitting the depth buffer if the depth buffer is GL_DEPTH24_STENCIL8. Changing the depth buffer format to GPU_DEPTH32F_STENCIL8 fixes the issue but only works if blitting the depth componnent. The stencil componnent still provoke issues when being copied.
2018-10-25 21:57:40 +02:00
bool GPU_depth_blitting_workaround(void)
{
return GG.depth_blitting_workaround;
}
Workbench: Fix workbench broken on some config due to usuned fb slot This seems to be a driver bug. Only windows + Radeon HD 7500M seems to be affected. Fix can be extended to more config if necessary.
2018-12-04 20:29:30 +01:00
bool GPU_unused_fb_slot_workaround(void)
{
return GG.unused_fb_slot_workaround;
}
Ghost Context Refactor https://developer.blender.org/D643 Separates graphics context creation from window code in Ghost so that they can vary separately.
2014-10-07 15:46:19 -05:00
void gpu_extensions_init(void)
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
{
OpenGL: make platform requirements explicit At context startup, make sure our assumptions about the OpenGL version are true. Should match since we set up the contexts... but this is what asserts are for, to check "should"s! Part of T49012
2016-09-30 19:10:30 -04:00
/* during 2.8 development each platform has its own OpenGL minimum requirements
OpenGL: add GLSL #version 330, drop 140 and 150 GL 3.3 is the new minimum. Compatibility profile for now, core profile eventually. During development, GL 3.0 (on Mesa) and 2.1 (on Mac) will still work. Part of T49012
2016-10-16 03:18:17 -04:00
* final 2.8 release will be unified on OpenGL 3.3 core profile, no required extensions
OpenGL: make platform requirements explicit At context startup, make sure our assumptions about the OpenGL version are true. Should match since we set up the contexts... but this is what asserts are for, to check "should"s! Part of T49012
2016-09-30 19:10:30 -04:00
* see developer.blender.org/T49012 for details
*/
OpenGL: add GLSL #version 330, drop 140 and 150 GL 3.3 is the new minimum. Compatibility profile for now, core profile eventually. During development, GL 3.0 (on Mesa) and 2.1 (on Mac) will still work. Part of T49012
2016-10-16 03:18:17 -04:00
BLI_assert(GLEW_VERSION_3_3);
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
GPUShader: Manually validate sampler count This happens on NVidia GPUs, using more textures than the maximum allowed by the gl will NOT trigger a linking issue (maybe because of bindless texture implementation?). So in this case we manually count the number of samplers per shader stage and compare it against the GL limit. We discard the shader if the sampler count is too high. This shows the user something is wrong with the shader.
2018-09-18 14:22:42 +02:00
glGetIntegerv(GL_MAX_TEXTURE_IMAGE_UNITS, &GG.maxtexturesfrag);
glGetIntegerv(GL_MAX_VERTEX_TEXTURE_IMAGE_UNITS, &GG.maxtexturesvert);
glGetIntegerv(GL_MAX_GEOMETRY_TEXTURE_IMAGE_UNITS, &GG.maxtexturesgeom);
glGetIntegerv(GL_MAX_COMBINED_TEXTURE_IMAGE_UNITS, &GG.maxtextures);
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
Add function to query maximum texture size. Also, make texture upload functions aware of this limit.
2013-04-12 17:56:07 +00:00
glGetIntegerv(GL_MAX_TEXTURE_SIZE, &GG.maxtexsize);
Fix T55888: Eevee: crash when shadow cube size is > 512px Note that this was only reported to happen on AMD GPU + windows.
2018-07-27 13:56:07 +02:00
glGetIntegerv(GL_MAX_ARRAY_TEXTURE_LAYERS, &GG.maxtexlayers);
World textures displaying for viewport in BI. This patch supports "Image or Movie" and "Environment map" types of world texture for the viewport. It supports: - "View", "AngMap" and "Equirectangular" types of mapping. - Different types of texture blending (according to BI world render). - Same color blending as when it lacked textures (but render via glsl). {F207734} {F207735} Example: {F275180} Original author: @valentin_b4w Regards, Alexander (Blend4Web Team). Reviewers: sergey, valentin_b4w, brecht, merwin Reviewed By: merwin Subscribers: campbellbarton, merwin, blueprintrandom, youle, a.romanov, yurikovelenov, AlexKowel, Evgeny_Rodygin Projects: #rendering, #opengl_gfx, #bf_blender:_next Differential Revision: https://developer.blender.org/D1414
2016-01-27 12:06:57 +03:00
glGetIntegerv(GL_MAX_CUBE_MAP_TEXTURE_SIZE, &GG.maxcubemapsize);
Add function to query maximum texture size. Also, make texture upload functions aware of this limit.
2013-04-12 17:56:07 +00:00
OpenGL: fix max texture anisotropy check Query max supported on init, use that to clamp user-set values.
2016-01-07 11:21:08 -05:00
if (GLEW_EXT_texture_filter_anisotropic)
glGetFloatv(GL_MAX_TEXTURE_MAX_ANISOTROPY_EXT, &GG.max_anisotropy);
else
GG.max_anisotropy = 1.0f;
DRW: Change UBOs binding logic. Use the same logic than textures. Also reset bindings only on shader changes.
2018-03-09 23:50:30 +01:00
glGetIntegerv(GL_MAX_FRAGMENT_UNIFORM_BLOCKS, &GG.maxubobinds);
Clay-Engine (merge clay-engine) Initial work by Clément Foucault with contributions from Dalai Felinto (mainly per-collection engine settings logic, and depsgraph iterator placeholder). This makes Blender require OpenGL 3.3. Which means Intel graphic card and OSX will break. Disable CLAY_ENGINE in CMake in those cases. This is a prototype render engine intended to help the design of real render engines. This is mainly an engine with enphasis in matcap and ambient occlusion. Implemented Features -------------------- * Clay Render Engine, following the new API, to be used as reference for future engines * A more complete Matcap customization with more options * Per-Collection render engine settings * New Ground Truth AO - not enabled Missing Features ---------------- * Finish object edit mode - Fix shaders to use new matrix - Fix artifacts when edge does off screen - Fix depth issue - Selection sillhouette - Mesh wires - Use mesh normals (for higher quality matcap) - Non-Mesh objects drawing - Widget drawing - Performance issues * Finish mesh edit mode - Derived-Mesh-less edit mode API (mesh_rende.c) * General edit mode - Per-collection edit mode settings * General engines - Per-collection engine settings (they are their, but they still need to be flushed by depsgraph, and used by the drawing code)
2017-02-07 11:20:15 +01:00
glGetIntegerv(GL_MAX_UNIFORM_BLOCK_SIZE, &GG.maxubosize);
GPU: Add safety check for max line width On some platform does not support line width > 1.0 and can even throw and error. Better check an at least display something rather than no lines at all.
2018-11-02 14:58:49 +01:00
glGetFloatv(GL_ALIASED_LINE_WIDTH_RANGE, GG.line_width_range);
Correct core profile way of getting color depth
2017-05-19 14:10:10 +02:00
#ifndef NDEBUG
GLint ret;
glBindFramebuffer(GL_FRAMEBUFFER, 0);
glGetFramebufferAttachmentParameteriv(GL_FRAMEBUFFER, GL_FRONT_LEFT, GL_FRAMEBUFFER_ATTACHMENT_OBJECT_TYPE, &ret);
/* We expect FRONT_LEFT to be the default buffer. */
Fix assert on framebuffer binding
2017-05-19 14:39:27 +02:00
BLI_assert(ret == GL_FRAMEBUFFER_DEFAULT);
Correct core profile way of getting color depth
2017-05-19 14:10:10 +02:00
#endif
GLint r, g, b;
glGetFramebufferAttachmentParameteriv(GL_FRAMEBUFFER, GL_FRONT_LEFT, GL_FRAMEBUFFER_ATTACHMENT_RED_SIZE, &r);
glGetFramebufferAttachmentParameteriv(GL_FRAMEBUFFER, GL_FRONT_LEFT, GL_FRAMEBUFFER_ATTACHMENT_GREEN_SIZE, &g);
glGetFramebufferAttachmentParameteriv(GL_FRAMEBUFFER, GL_FRONT_LEFT, GL_FRAMEBUFFER_ATTACHMENT_BLUE_SIZE, &b);
GG.colordepth = r + g + b; /* Assumes same depth for RGB. */
whitespace only, no functional change mixed tabs/spaces --> tabs.
2011-04-21 13:11:51 +00:00
GPUShader: Manually validate sampler count This happens on NVidia GPUs, using more textures than the maximum allowed by the gl will NOT trigger a linking issue (maybe because of bindless texture implementation?). So in this case we manually count the number of samplers per shader stage and compare it against the GL limit. We discard the shader if the sampler count is too high. This shows the user something is wrong with the shader.
2018-09-18 14:22:42 +02:00
glGetIntegerv(GL_MAX_COLOR_TEXTURE_SAMPLES, &GG.samples_color_texture_max);
Support for multi-sample off-screen buffers Replaces much slower manual accumulation buffer which simply did multiple renders. Needs OpenGL3.2, otherwise multi-sample is disabled.
2015-10-14 10:33:30 +11:00
OpenGL: GLSL always supported In gpu lib: - GPU_glsl_support() always returns true - internal cleanup & comments Outside gpu lib: - remove check from various code, remove the “else” path - sprinkled a few C99-isms We can remove GPU_glsl_support() when BGE stops calling it.
2015-11-25 11:49:46 -05:00
const char *vendor = (const char *)glGetString(GL_VENDOR);
const char *renderer = (const char *)glGetString(GL_RENDERER);
const char *version = (const char *)glGetString(GL_VERSION);
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
GPU: Fix missing case of new radeon on xorg. This was causing issue with shader compilation.
2018-06-04 15:14:52 +02:00
if (strstr(vendor, "ATI") || strstr(vendor, "AMD")) {
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
GG.device = GPU_DEVICE_ATI;
GG.driver = GPU_DRIVER_OFFICIAL;
GPU: Add workarounds for buggy glBlitFramebuffer function on macOS + Radeon When calling glBlitFramebuffer on most (if not all) mac that have a GPU from the Radeon Pro series, the data is not properly copied and only a subset of the pixels are correctly copied. This only happens when blitting the depth buffer if the depth buffer is GL_DEPTH24_STENCIL8. Changing the depth buffer format to GPU_DEPTH32F_STENCIL8 fixes the issue but only works if blitting the depth componnent. The stencil componnent still provoke issues when being copied.
2018-10-25 21:57:40 +02:00
Workbench: Fix workbench broken on some config due to usuned fb slot This seems to be a driver bug. Only windows + Radeon HD 7500M seems to be affected. Fix can be extended to more config if necessary.
2018-12-04 20:29:30 +01:00
#ifdef _WIN32
Fix T56362: sub surface scattering does not work on eevee on radeon hd 7600 series
2018-12-05 21:22:12 -02:00
if (strstr(version, "4.5.13399") ||
GPU: Add another driver to the list of unused_fb_slot_workaround I start to think that an automatic detection would be a better solution.
2018-12-25 14:36:22 -02:00
strstr(version, "4.5.13417") ||
strstr(version, "4.5.13422"))
Workbench: Fix workbench broken on some config due to usuned fb slot This seems to be a driver bug. Only windows + Radeon HD 7500M seems to be affected. Fix can be extended to more config if necessary.
2018-12-04 20:29:30 +01:00
{
GPU Extensions: unused_fb_slot_workaround: Compare only the driver. Dozens of renderes are included.
2018-12-06 10:05:42 -02:00
/* The renderers include:
* Mobility Radeon HD 5000;
* Radeon HD 7500M;
* Radeon HD 7570M;
* Radeon HD 7600M;
* And many others... */
Workbench: Fix workbench broken on some config due to usuned fb slot This seems to be a driver bug. Only windows + Radeon HD 7500M seems to be affected. Fix can be extended to more config if necessary.
2018-12-04 20:29:30 +01:00
GG.unused_fb_slot_workaround = true;
}
#endif
GPU: Add workarounds for buggy glBlitFramebuffer function on macOS + Radeon When calling glBlitFramebuffer on most (if not all) mac that have a GPU from the Radeon Pro series, the data is not properly copied and only a subset of the pixels are correctly copied. This only happens when blitting the depth buffer if the depth buffer is GL_DEPTH24_STENCIL8. Changing the depth buffer format to GPU_DEPTH32F_STENCIL8 fixes the issue but only works if blitting the depth componnent. The stencil componnent still provoke issues when being copied.
2018-10-25 21:57:40 +02:00
#if defined(__APPLE__)
GPU: Add AMD Radeon RX series to macos blitting workaround list Fixes T55987
2018-11-28 13:19:04 +01:00
if (strstr(renderer, "AMD Radeon Pro") ||
strstr(renderer, "AMD Radeon R9") ||
strstr(renderer, "AMD Radeon RX"))
{
GPU: Add workarounds for buggy glBlitFramebuffer function on macOS + Radeon When calling glBlitFramebuffer on most (if not all) mac that have a GPU from the Radeon Pro series, the data is not properly copied and only a subset of the pixels are correctly copied. This only happens when blitting the depth buffer if the depth buffer is GL_DEPTH24_STENCIL8. Changing the depth buffer format to GPU_DEPTH32F_STENCIL8 fixes the issue but only works if blitting the depth componnent. The stencil componnent still provoke issues when being copied.
2018-10-25 21:57:40 +02:00
GG.depth_blitting_workaround = true;
}
#endif
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (strstr(vendor, "NVIDIA")) {
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
GG.device = GPU_DEVICE_NVIDIA;
GG.driver = GPU_DRIVER_OFFICIAL;
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (strstr(vendor, "Intel") ||
Cleanup: style, gpu module
2016-01-10 07:12:10 +11:00
/* src/mesa/drivers/dri/intel/intel_context.c */
strstr(renderer, "Mesa DRI Intel") ||
strstr(renderer, "Mesa DRI Mobile Intel"))
{
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
GG.device = GPU_DEVICE_INTEL;
GG.driver = GPU_DRIVER_OFFICIAL;
Fix T57455: Laggy, freezing UI with Linux and Intel UHD 620 Seems like a driver bug but doing glFlush() before these calls fixes it.
2018-12-02 01:57:22 +01:00
GPU: Improve detection of intels UHD graphics
2018-12-05 04:22:32 +01:00
if (strstr(renderer, "UHD Graphics") ||
GPU: Add Intel HD Graphics 530 to the list of buggy intel gpus
2018-12-17 08:56:38 +01:00
/* Not UHD but affected by the same bugs. */
strstr(renderer, "HD Graphics 530") ||
GPU: Improve detection of intels UHD graphics
2018-12-05 04:22:32 +01:00
strstr(renderer, "Kaby Lake GT2"))
{
Fix T57455: Laggy, freezing UI with Linux and Intel UHD 620 Seems like a driver bug but doing glFlush() before these calls fixes it.
2018-12-02 01:57:22 +01:00
GG.device |= GPU_DEVICE_INTEL_UHD;
}
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
}
GPU: Consider latest Gallium driver an official ATI/AMD This will make triple buffer used by default for such configuration. Ideally we would switch to triple buffer on all platforms, but let's do it in 2.8 branch and don't open can of worms in master now. This should solve issues like T49945.
2016-11-22 11:38:27 +01:00
else if ((strstr(renderer, "Mesa DRI R")) ||
GPU: Fix missing case of new radeon on xorg. This was causing issue with shader compilation.
2018-06-04 15:14:52 +02:00
(strstr(renderer, "Radeon") && strstr(vendor, "X.Org")) ||
GPU: Consider latest Gallium driver an official ATI/AMD This will make triple buffer used by default for such configuration. Ideally we would switch to triple buffer on all platforms, but let's do it in 2.8 branch and don't open can of worms in master now. This should solve issues like T49945.
2016-11-22 11:38:27 +01:00
(strstr(renderer, "Gallium ") && strstr(renderer, " on ATI ")) ||
(strstr(renderer, "Gallium ") && strstr(renderer, " on AMD ")))
{
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
GG.device = GPU_DEVICE_ATI;
GG.driver = GPU_DRIVER_OPENSOURCE;
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (strstr(renderer, "Nouveau") || strstr(vendor, "nouveau")) {
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
GG.device = GPU_DEVICE_NVIDIA;
GG.driver = GPU_DRIVER_OPENSOURCE;
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (strstr(vendor, "Mesa")) {
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
GG.device = GPU_DEVICE_SOFTWARE;
GG.driver = GPU_DRIVER_SOFTWARE;
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (strstr(vendor, "Microsoft")) {
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
GG.device = GPU_DEVICE_SOFTWARE;
GG.driver = GPU_DRIVER_SOFTWARE;
}
style cleanup: follow style guide for formatting of if/for/while loops, and else if's
2012-03-24 06:38:07 +00:00
else if (strstr(renderer, "Apple Software Renderer")) {
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
GG.device = GPU_DEVICE_SOFTWARE;
GG.driver = GPU_DRIVER_SOFTWARE;
}
Recognize llvmpipe renderer as software OpenGL
2019-01-04 12:41:35 +01:00
else if (strstr(renderer, "llvmpipe")) {
GG.device = GPU_DEVICE_SOFTWARE;
GG.driver = GPU_DRIVER_SOFTWARE;
}
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
else {
GPU: Fix missing case of new radeon on xorg. This was causing issue with shader compilation.
2018-06-04 15:14:52 +02:00
printf("Warning: Could not find a matching GPU name. Things may not behave as expected.\n");
Provide more information on unknown GPU
2019-01-04 12:50:05 +01:00
printf("Detected OpenGL configuration:\n");
printf("Vendor: %s\n", vendor);
printf("Renderer: %s\n", renderer);
2.5: automatic draw method now uses overlap for Intel on all platforms.
2010-07-16 10:13:04 +00:00
GG.device = GPU_DEVICE_ANY;
GG.driver = GPU_DRIVER_ANY;
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
}
#ifdef _WIN32
GG.os = GPU_OS_WIN;
code cleanup: duplicate checks and double assignments.
2012-03-08 04:38:35 +00:00
#elif defined(__APPLE__)
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
GG.os = GPU_OS_MAC;
code cleanup: duplicate checks and double assignments.
2012-03-08 04:38:35 +00:00
#else
GG.os = GPU_OS_UNIX;
WM Draw Methods now has a new option Automatic (default). This will set the draw method to triple buffer or overlap depending on the configuration. Ideally I could get all cases working well with triple buffer but it's hard in practice. At the moment there are two cases that use overlap instead: * opensource ATI drives on linux * windows software renderer Also added a utility function to check GPU device/os/driver.
2010-01-31 23:45:51 +00:00
#endif
OpenGL: implemenation of fixed function lighting as per pixel GLSL shaders. The code is still unused, but the intention is to use this to solve the double sided lighting problem on NVidia, and to make the materials work on OpenGL ES 2.0 eventually. The code works and matches the fixed function lighting pretty much exactly, but still needs optimizations. The actual integration in object draw will be committed later when more fixing & testing, there's lots of different combinations and unclear OpenGL state here.
2013-02-26 00:49:42 +00:00
GPU: Fix Issue with recursive downsample and Intel HDXXX This is caused by a driver bug that prevent us from rendering to (or even binding) a texture mip level that is below GL_TEXTURE_MAX_LEVEL of the target texture. This is fine in most drivers (and legal AFAIK) but not on thoses Intels HDXXX + Windows. As a fix we just put GL_TEXTURE_MAX_LEVEL lower (which is illegal because it is undefined behaviour), but in practice it works ok and does not trigger any warnings or errors. This commit fixes most of the problems encountered on these GPUs (T56668).
2018-10-22 12:58:11 +02:00
gpu_detect_mip_render_workaround();
Fix T40498 invalid textures flickering. Issue here is most likely sampler uniforms and textures not being updated properly when zero binding is created. Solution for now is to allow zero binding but when this happens use sexy pink invalid texture instead :p.
2014-06-10 01:36:53 +03:00
BKE: Add "--debug-gpu-force-workarounds" to force gpu workarounds This is nice to test workarounds on other configs that may benefits from the existing workarounds.
2018-12-05 03:05:39 +01:00
if (G.debug & G_DEBUG_GPU_FORCE_WORKAROUNDS) {
GPU: Print opengl infos if using --debug-gpu-force-workarounds
2018-12-05 04:03:40 +01:00
printf("\n");
printf("GPU: Bypassing workaround detection.\n");
Cleanup: Spelling
2019-01-04 12:51:10 +01:00
printf("GPU: OpenGL identification strings\n");
GPU: Print opengl infos if using --debug-gpu-force-workarounds
2018-12-05 04:03:40 +01:00
printf("GPU: vendor: %s\n", vendor);
printf("GPU: renderer: %s\n", renderer);
printf("GPU: version: %s\n\n", version);
BKE: Add "--debug-gpu-force-workarounds" to force gpu workarounds This is nice to test workarounds on other configs that may benefits from the existing workarounds.
2018-12-05 03:05:39 +01:00
GG.mip_render_workaround = true;
GG.depth_blitting_workaround = true;
GG.unused_fb_slot_workaround = true;
}
Potential fix for T43987, ambient occlusion different between offscreen and on screen rendering. Aaaaah, the beauty of driver implementations of OpenGL! Turns out the problem here is that drivers calculate df/dy differently in some cases (probably because OpenGL counts y reverse to how the window system does, so drivers can get confused). Fixed this for the ATI case based on info we have so far, there's also the Intel case which will be handled separately (missing info on Intel's renderer string etc). Unfortunately we can't really fix this for the general case so we'll have to haldle cases as they come in our tracker and by adding silly string comparisons in our GPU initialization module <sigh>.
2015-03-30 14:14:32 +02:00
/* df/dy calculation factors, those are dependent on driver */
if ((strstr(vendor, "ATI") && strstr(version, "3.3.10750"))) {
GG.dfdyfactors[0] = 1.0;
GG.dfdyfactors[1] = -1.0;
}
Cleanup: style, gpu module
2016-01-10 07:12:10 +11:00
else if ((GG.device == GPU_DEVICE_INTEL) && (GG.os == GPU_OS_WIN) &&
(strstr(version, "4.0.0 - Build 10.18.10.3308") ||
strstr(version, "4.0.0 - Build 9.18.10.3186") ||
strstr(version, "4.0.0 - Build 9.18.10.3165") ||
strstr(version, "3.1.0 - Build 9.17.10.3347") ||
strstr(version, "3.1.0 - Build 9.17.10.4101") ||
strstr(version, "3.3.0 - Build 8.15.10.2618")))
Final solution for Intel card Ambient Occlusion in T43987. Forgot to account for offscreen case in addition to compositing
2015-05-18 22:05:38 +02:00
{
GG.dfdyfactors[0] = -1.0;
GG.dfdyfactors[1] = 1.0;
Potential fix for T43987, ambient occlusion different between offscreen and on screen rendering. Aaaaah, the beauty of driver implementations of OpenGL! Turns out the problem here is that drivers calculate df/dy differently in some cases (probably because OpenGL counts y reverse to how the window system does, so drivers can get confused). Fixed this for the ATI case based on info we have so far, there's also the Intel case which will be handled separately (missing info on Intel's renderer string etc). Unfortunately we can't really fix this for the general case so we'll have to haldle cases as they come in our tracker and by adding silly string comparisons in our GPU initialization module <sigh>.
2015-03-30 14:14:32 +02:00
}
else {
GG.dfdyfactors[0] = 1.0;
GG.dfdyfactors[1] = 1.0;
}
Fix T40498 invalid textures flickering. Issue here is most likely sampler uniforms and textures not being updated properly when zero binding is created. Solution for now is to allow zero binding but when this happens use sexy pink invalid texture instead :p.
2014-06-10 01:36:53 +03:00
GPU_invalid_tex_init();
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
}
Ghost Context Refactor https://developer.blender.org/D643 Separates graphics context creation from window code in Ghost so that they can vary separately.
2014-10-07 15:46:19 -05:00
void gpu_extensions_exit(void)
Cleaning up the GPU_extensions_init/exit() code a bit to keep the Blenderplayer from crashing on exit and restart.
2011-12-23 07:10:01 +00:00
{
Fix T40498 invalid textures flickering. Issue here is most likely sampler uniforms and textures not being updated properly when zero binding is created. Solution for now is to allow zero binding but when this happens use sexy pink invalid texture instead :p.
2014-06-10 01:36:53 +03:00
GPU_invalid_tex_free();
Cleaning up the GPU_extensions_init/exit() code a bit to keep the Blenderplayer from crashing on exit and restart.
2011-12-23 07:10:01 +00:00
}
OpenGL: new GPU_legacy_support() function Is current context compatible with legacy GL (version 2.1)? My earlier approach -- checking for GLEW_ARB_compatibility -- was not enough. This should always return true if we set our GL context up properly. It will return false when we switch to core profile.
2015-12-06 18:00:22 -05:00
warning cleanup. - fix mistake with grease pencil UI (&& was intended but & used). - use (void) rather then () across _all_ blenders code. - a few minor edits, don't shadow stack variables in roll calculation & avoid running memset() for VBO vertex map.
2011-02-13 15:02:21 +00:00
int GPU_color_depth(void)
Fix #21349: triple buffer drawing doesn't work well with thousands of colors setting on Mac, just disabled it in that case.
2010-04-05 10:25:40 +00:00
{
whitespace only, no functional change mixed tabs/spaces --> tabs.
2011-04-21 13:11:51 +00:00
return GG.colordepth;
Fix #21349: triple buffer drawing doesn't work well with thousands of colors setting on Mac, just disabled it in that case.
2010-04-05 10:25:40 +00:00
}
OpenGL: split off framebuffer, shader and texture code into separate files.
2015-12-06 21:20:19 +01:00
bool GPU_mem_stats_supported(void)
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
{
OpenGL: minor cleanup & TODO Just clearing out some old git stashes. No point getting old GL light model to work on old Intel GPUs (both are obsolete for 2.8)
2017-04-15 19:43:56 -04:00
return (GLEW_NVX_gpu_memory_info || GLEW_ATI_meminfo) && (G.debug & G_DEBUG_GPU_MEM);
Merge of first part of changes from the apricot branch, especially the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
2008-09-04 20:51:28 +00:00
}
OpenGL: split off framebuffer, shader and texture code into separate files.
2015-12-06 21:20:19 +01:00
void GPU_mem_stats_get(int *totalmem, int *freemem)
Viewport compositing - first code This commit introduces a few ready made effects for the 3D viewport and OpenGL rendering. Included effects are Depth of Field, accessible from camera view and screen space ambient occlusion. Those effects can be turned on and tweaked from the shading panel in the 3D viewport. Off screen rendering will use the settings of the current camera. WIP documentation can be found here: http://wiki.blender.org/index.php/User:Psy-Fi/Framebuffer_Post-processing
2015-02-12 18:54:41 +01:00
{
OpenGL: minor cleanup & TODO Just clearing out some old git stashes. No point getting old GL light model to work on old Intel GPUs (both are obsolete for 2.8)
2017-04-15 19:43:56 -04:00
/* TODO(merwin): use Apple's platform API to get this info */
OpenGL: split off framebuffer, shader and texture code into separate files.
2015-12-06 21:20:19 +01:00
if (GLEW_NVX_gpu_memory_info) {
/* returned value in Kb */
glGetIntegerv(GL_GPU_MEMORY_INFO_TOTAL_AVAILABLE_MEMORY_NVX, totalmem);
Depth of field high quality: A new checkbox "High quality" is provided in camera settings to enable this. This creates a depth of field that is much closer to the rendered result and even supports aperture blades in the effect, but it's more expensive too. There are optimizations to do here since the technique is very fill rate heavy. People, be careful, this -can- lock up your screen if depth of field blurring is too extreme. Technical details: This uses geometry shaders + instancing and is an adaptation of techniques gathered from http://bartwronski.com/2014/04/07/bokeh-depth-of-field-going-insane- http://advances.realtimerendering.com/s2011/SousaSchulzKazyan%20- %20in%20Real-Time%20Rendering%20Course).ppt TODOs: * Support dithering to minimize banding. * Optimize fill rate in geometry shader.
2015-03-03 17:47:31 +01:00
OpenGL: split off framebuffer, shader and texture code into separate files.
2015-12-06 21:20:19 +01:00
glGetIntegerv(GL_GPU_MEMORY_INFO_CURRENT_AVAILABLE_VIDMEM_NVX, freemem);
Depth of field high quality: A new checkbox "High quality" is provided in camera settings to enable this. This creates a depth of field that is much closer to the rendered result and even supports aperture blades in the effect, but it's more expensive too. There are optimizations to do here since the technique is very fill rate heavy. People, be careful, this -can- lock up your screen if depth of field blurring is too extreme. Technical details: This uses geometry shaders + instancing and is an adaptation of techniques gathered from http://bartwronski.com/2014/04/07/bokeh-depth-of-field-going-insane- http://advances.realtimerendering.com/s2011/SousaSchulzKazyan%20- %20in%20Real-Time%20Rendering%20Course).ppt TODOs: * Support dithering to minimize banding. * Optimize fill rate in geometry shader.
2015-03-03 17:47:31 +01:00
}
OpenGL: split off framebuffer, shader and texture code into separate files.
2015-12-06 21:20:19 +01:00
else if (GLEW_ATI_meminfo) {
int stats[4];
Viewport compositing - first code This commit introduces a few ready made effects for the 3D viewport and OpenGL rendering. Included effects are Depth of Field, accessible from camera view and screen space ambient occlusion. Those effects can be turned on and tweaked from the shading panel in the 3D viewport. Off screen rendering will use the settings of the current camera. WIP documentation can be found here: http://wiki.blender.org/index.php/User:Psy-Fi/Framebuffer_Post-processing
2015-02-12 18:54:41 +01:00
OpenGL: split off framebuffer, shader and texture code into separate files.
2015-12-06 21:20:19 +01:00
glGetIntegerv(GL_TEXTURE_FREE_MEMORY_ATI, stats);
*freemem = stats[0];
*totalmem = 0;
Add debug option --debug-gpumem to show GPU memory used in status bar. Only used in ATIs and NVIDIAs. Used extensions are: https://www.opengl.org/registry/specs/ATI/meminfo.txt http://developer.download.nvidia.com/opengl/specs/ If you read the documentation, the numbers are not supposed to be exact and also depend on the time when the call is made. The numbers can also change quite quickly. It's only meant to give a rough measure of what is going on.
2015-04-24 14:11:05 +02:00
}
else {
*totalmem = 0;
*freemem = 0;
}
}
Copy Permalink