kwin/lanczosfilter.cpp
Martin Gräßlin 1f52366fc1 Disable ARB Lanczos Shader for Intel GPUs
For Sandybridge Lanczos Shader is enabled, but only GLSL is working
so the ARB variant has to be disabled for Intel GPUs.

BUG: 301729
FIXED-IN: 4.9.0
REVIEW: 105233
2012-06-12 22:53:36 +02:00

707 lines
27 KiB
C++

/********************************************************************
KWin - the KDE window manager
This file is part of the KDE project.
Copyright (C) 2010 by Fredrik Höglund <fredrik@kde.org>
Copyright (C) 2010 Martin Gräßlin <kde@martin-graesslin.com>
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 of the License, or
(at your option) any later version.
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, see <http://www.gnu.org/licenses/>.
*********************************************************************/
#include "lanczosfilter.h"
#include "effects.h"
#include "options.h"
#include <kwinglutils.h>
#include <kwinglplatform.h>
#include <kwineffects.h>
#include <KDE/KGlobalSettings>
#include <qmath.h>
#include <cmath>
namespace KWin
{
LanczosFilter::LanczosFilter(QObject* parent)
: QObject(parent)
, m_offscreenTex(0)
, m_offscreenTarget(0)
, m_shader(0)
, m_inited(false)
{
}
LanczosFilter::~LanczosFilter()
{
delete m_offscreenTarget;
delete m_offscreenTex;
}
void LanczosFilter::init()
{
if (m_inited)
return;
m_inited = true;
const bool force = (qstrcmp(qgetenv("KWIN_FORCE_LANCZOS"), "1") == 0);
if (force) {
kWarning(1212) << "Lanczos Filter forced on by environment variable";
}
if (!force && options->glSmoothScale() != 2)
return; // disabled by config
// The lanczos filter is reported to be broken with the Intel driver and Mesa 7.10
GLPlatform *gl = GLPlatform::instance();
if (!force && gl->driver() == Driver_Intel && gl->mesaVersion() >= kVersionNumber(7, 10) && gl->chipClass() < SandyBridge)
return;
// With fglrx the ARB Shader crashes KWin (see Bug #270818 and #286795) and GLSL Shaders are not functional
if (!force && gl->driver() == Driver_Catalyst) {
return;
}
m_shader = new LanczosShader(this);
if (!m_shader->init()) {
delete m_shader;
m_shader = 0;
}
}
void LanczosFilter::updateOffscreenSurfaces()
{
int w = displayWidth();
int h = displayHeight();
if (!GLTexture::NPOTTextureSupported()) {
w = nearestPowerOfTwo(w);
h = nearestPowerOfTwo(h);
}
if (!m_offscreenTex || m_offscreenTex->width() != w || m_offscreenTex->height() != h) {
if (m_offscreenTex) {
delete m_offscreenTex;
delete m_offscreenTarget;
}
m_offscreenTex = new GLTexture(w, h);
m_offscreenTex->setFilter(GL_LINEAR);
m_offscreenTex->setWrapMode(GL_CLAMP_TO_EDGE);
m_offscreenTarget = new GLRenderTarget(*m_offscreenTex);
}
}
static float sinc(float x)
{
return std::sin(x * M_PI) / (x * M_PI);
}
static float lanczos(float x, float a)
{
if (qFuzzyCompare(x + 1.0, 1.0))
return 1.0;
if (qAbs(x) >= a)
return 0.0;
return sinc(x) * sinc(x / a);
}
void LanczosShader::createKernel(float delta, int *size)
{
const float a = 2.0;
// The two outermost samples always fall at points where the lanczos
// function returns 0, so we'll skip them.
const int sampleCount = qBound(3, qCeil(delta * a) * 2 + 1 - 2, 29);
const int center = sampleCount / 2;
const int kernelSize = center + 1;
const float factor = 1.0 / delta;
QVector<float> values(kernelSize);
float sum = 0;
for (int i = 0; i < kernelSize; i++) {
const float val = lanczos(i * factor, a);
sum += i > 0 ? val * 2 : val;
values[i] = val;
}
memset(m_kernel, 0, 16 * sizeof(QVector4D));
// Normalize the kernel
for (int i = 0; i < kernelSize; i++) {
const float val = values[i] / sum;
m_kernel[i] = QVector4D(val, val, val, val);
}
*size = kernelSize;
}
void LanczosShader::createOffsets(int count, float width, Qt::Orientation direction)
{
memset(m_offsets, 0, 16 * sizeof(QVector2D));
for (int i = 0; i < count; i++) {
m_offsets[i] = (direction == Qt::Horizontal) ?
QVector2D(i / width, 0) : QVector2D(0, i / width);
}
}
void LanczosFilter::performPaint(EffectWindowImpl* w, int mask, QRegion region, WindowPaintData& data)
{
if (effects->compositingType() == KWin::OpenGLCompositing && (data.xScale < 0.9 || data.yScale < 0.9) &&
KGlobalSettings::graphicEffectsLevel() & KGlobalSettings::SimpleAnimationEffects) {
if (!m_inited)
init();
const QRect screenRect = Workspace::self()->clientArea(ScreenArea, w->screen(), w->desktop());
// window geometry may not be bigger than screen geometry to fit into the FBO
if (m_shader && w->width() <= screenRect.width() && w->height() <= screenRect.height()) {
double left = 0;
double top = 0;
double right = w->width();
double bottom = w->height();
foreach (const WindowQuad & quad, data.quads) {
// we need this loop to include the decoration padding
left = qMin(left, quad.left());
top = qMin(top, quad.top());
right = qMax(right, quad.right());
bottom = qMax(bottom, quad.bottom());
}
double width = right - left;
double height = bottom - top;
if (width > screenRect.width() || height > screenRect.height()) {
// window with padding does not fit into the framebuffer
// so cut of the shadow
left = 0;
top = 0;
width = w->width();
height = w->height();
}
int tx = data.xTranslate + w->x() + left * data.xScale;
int ty = data.yTranslate + w->y() + top * data.yScale;
int tw = width * data.xScale;
int th = height * data.yScale;
const QRect textureRect(tx, ty, tw, th);
const bool hardwareClipping = !(QRegion(textureRect)-region).isEmpty();
int sw = width;
int sh = height;
GLTexture *cachedTexture = static_cast< GLTexture*>(w->data(LanczosCacheRole).value<void*>());
if (cachedTexture) {
if (cachedTexture->width() == tw && cachedTexture->height() == th) {
cachedTexture->bind();
if (hardwareClipping) {
glEnable(GL_SCISSOR_TEST);
}
if (ShaderManager::instance()->isValid()) {
glEnable(GL_BLEND);
glBlendFunc(GL_SRC_ALPHA, GL_ONE_MINUS_SRC_ALPHA);
const float rgb = data.brightness * data.opacity;
const float a = data.opacity;
GLShader *shader = ShaderManager::instance()->pushShader(ShaderManager::SimpleShader);
shader->setUniform(GLShader::Offset, QVector2D(0, 0));
shader->setUniform(GLShader::ModulationConstant, QVector4D(rgb, rgb, rgb, a));
shader->setUniform(GLShader::Saturation, data.saturation);
shader->setUniform(GLShader::AlphaToOne, 0);
cachedTexture->render(region, textureRect, hardwareClipping);
ShaderManager::instance()->popShader();
glDisable(GL_BLEND);
} else {
prepareRenderStates(cachedTexture, data.opacity, data.brightness, data.saturation);
cachedTexture->render(region, textureRect, hardwareClipping);
restoreRenderStates(cachedTexture, data.opacity, data.brightness, data.saturation);
}
if (hardwareClipping) {
glDisable(GL_SCISSOR_TEST);
}
cachedTexture->unbind();
m_timer.start(5000, this);
return;
} else {
// offscreen texture not matching - delete
delete cachedTexture;
cachedTexture = 0;
w->setData(LanczosCacheRole, QVariant());
}
}
WindowPaintData thumbData = data;
thumbData.xScale = 1.0;
thumbData.yScale = 1.0;
thumbData.xTranslate = -w->x() - left;
thumbData.yTranslate = -w->y() - top;
thumbData.brightness = 1.0;
thumbData.opacity = 1.0;
thumbData.saturation = 1.0;
// Bind the offscreen FBO and draw the window on it unscaled
updateOffscreenSurfaces();
GLRenderTarget::pushRenderTarget(m_offscreenTarget);
glClearColor(0.0, 0.0, 0.0, 0.0);
glClear(GL_COLOR_BUFFER_BIT);
w->sceneWindow()->performPaint(mask, infiniteRegion(), thumbData);
// Create a scratch texture and copy the rendered window into it
GLTexture tex(sw, sh);
tex.setFilter(GL_LINEAR);
tex.setWrapMode(GL_CLAMP_TO_EDGE);
tex.bind();
glCopyTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, 0, m_offscreenTex->height() - sh, sw, sh);
// Set up the shader for horizontal scaling
float dx = sw / float(tw);
int kernelSize;
m_shader->createKernel(dx, &kernelSize);
m_shader->createOffsets(kernelSize, sw, Qt::Horizontal);
m_shader->bind();
m_shader->setUniforms();
// Draw the window back into the FBO, this time scaled horizontally
glClear(GL_COLOR_BUFFER_BIT);
QVector<float> verts;
QVector<float> texCoords;
verts.reserve(12);
texCoords.reserve(12);
texCoords << 1.0 << 0.0; verts << tw << 0.0; // Top right
texCoords << 0.0 << 0.0; verts << 0.0 << 0.0; // Top left
texCoords << 0.0 << 1.0; verts << 0.0 << sh; // Bottom left
texCoords << 0.0 << 1.0; verts << 0.0 << sh; // Bottom left
texCoords << 1.0 << 1.0; verts << tw << sh; // Bottom right
texCoords << 1.0 << 0.0; verts << tw << 0.0; // Top right
GLVertexBuffer *vbo = GLVertexBuffer::streamingBuffer();
vbo->reset();
vbo->setData(6, 2, verts.constData(), texCoords.constData());
vbo->render(GL_TRIANGLES);
// At this point we don't need the scratch texture anymore
tex.unbind();
tex.discard();
// create scratch texture for second rendering pass
GLTexture tex2(tw, sh);
tex2.setFilter(GL_LINEAR);
tex2.setWrapMode(GL_CLAMP_TO_EDGE);
tex2.bind();
glCopyTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, 0, m_offscreenTex->height() - sh, tw, sh);
// Set up the shader for vertical scaling
float dy = sh / float(th);
m_shader->createKernel(dy, &kernelSize);
m_shader->createOffsets(kernelSize, m_offscreenTex->height(), Qt::Vertical);
m_shader->setUniforms();
// Now draw the horizontally scaled window in the FBO at the right
// coordinates on the screen, while scaling it vertically and blending it.
glClear(GL_COLOR_BUFFER_BIT);
verts.clear();
verts << tw << 0.0; // Top right
verts << 0.0 << 0.0; // Top left
verts << 0.0 << th; // Bottom left
verts << 0.0 << th; // Bottom left
verts << tw << th; // Bottom right
verts << tw << 0.0; // Top right
vbo->setData(6, 2, verts.constData(), texCoords.constData());
vbo->render(GL_TRIANGLES);
tex2.unbind();
tex2.discard();
m_shader->unbind();
// create cache texture
GLTexture *cache = new GLTexture(tw, th);
cache->setFilter(GL_LINEAR);
cache->setWrapMode(GL_CLAMP_TO_EDGE);
cache->bind();
glCopyTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, 0, m_offscreenTex->height() - th, tw, th);
GLRenderTarget::popRenderTarget();
if (hardwareClipping) {
glEnable(GL_SCISSOR_TEST);
}
if (ShaderManager::instance()->isValid()) {
glEnable(GL_BLEND);
glBlendFunc(GL_SRC_ALPHA, GL_ONE_MINUS_SRC_ALPHA);
const float rgb = data.brightness * data.opacity;
const float a = data.opacity;
GLShader *shader = ShaderManager::instance()->pushShader(ShaderManager::SimpleShader);
shader->setUniform(GLShader::Offset, QVector2D(0, 0));
shader->setUniform(GLShader::ModulationConstant, QVector4D(rgb, rgb, rgb, a));
shader->setUniform(GLShader::Saturation, data.saturation);
shader->setUniform(GLShader::AlphaToOne, 0);
cache->render(region, textureRect, hardwareClipping);
ShaderManager::instance()->popShader();
glDisable(GL_BLEND);
} else {
prepareRenderStates(cache, data.opacity, data.brightness, data.saturation);
cache->render(region, textureRect, hardwareClipping);
restoreRenderStates(cache, data.opacity, data.brightness, data.saturation);
}
if (hardwareClipping) {
glDisable(GL_SCISSOR_TEST);
}
cache->unbind();
w->setData(LanczosCacheRole, QVariant::fromValue(static_cast<void*>(cache)));
// Delete the offscreen surface after 5 seconds
m_timer.start(5000, this);
return;
}
} // if ( effects->compositingType() == KWin::OpenGLCompositing )
w->sceneWindow()->performPaint(mask, region, data);
} // End of function
void LanczosFilter::timerEvent(QTimerEvent *event)
{
if (event->timerId() == m_timer.timerId()) {
m_timer.stop();
delete m_offscreenTarget;
delete m_offscreenTex;
m_offscreenTarget = 0;
m_offscreenTex = 0;
foreach (EffectWindow * w, effects->stackingOrder()) {
QVariant cachedTextureVariant = w->data(LanczosCacheRole);
if (cachedTextureVariant.isValid()) {
GLTexture *cachedTexture = static_cast< GLTexture*>(cachedTextureVariant.value<void*>());
delete cachedTexture;
cachedTexture = 0;
w->setData(LanczosCacheRole, QVariant());
}
}
}
}
void LanczosFilter::prepareRenderStates(GLTexture* tex, double opacity, double brightness, double saturation)
{
#ifdef KWIN_HAVE_OPENGLES
Q_UNUSED(tex)
Q_UNUSED(opacity)
Q_UNUSED(brightness)
Q_UNUSED(saturation)
#else
const bool alpha = true;
// setup blending of transparent windows
glPushAttrib(GL_ENABLE_BIT);
glEnable(GL_BLEND);
glBlendFunc(GL_ONE, GL_ONE_MINUS_SRC_ALPHA);
if (saturation != 1.0 && tex->saturationSupported()) {
// First we need to get the color from [0; 1] range to [0.5; 1] range
glActiveTexture(GL_TEXTURE0);
glTexEnvi(GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_COMBINE);
glTexEnvi(GL_TEXTURE_ENV, GL_COMBINE_RGB, GL_INTERPOLATE);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE0_RGB, GL_TEXTURE);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND0_RGB, GL_SRC_COLOR);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE1_RGB, GL_CONSTANT);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND1_RGB, GL_SRC_COLOR);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE2_RGB, GL_CONSTANT);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND2_RGB, GL_SRC_ALPHA);
const float scale_constant[] = { 1.0, 1.0, 1.0, 0.5};
glTexEnvfv(GL_TEXTURE_ENV, GL_TEXTURE_ENV_COLOR, scale_constant);
tex->bind();
// Then we take dot product of the result of previous pass and
// saturation_constant. This gives us completely unsaturated
// (greyscale) image
// Note that both operands have to be in range [0.5; 1] since opengl
// automatically substracts 0.5 from them
glActiveTexture(GL_TEXTURE1);
float saturation_constant[] = { 0.5 + 0.5 * 0.30, 0.5 + 0.5 * 0.59, 0.5 + 0.5 * 0.11, saturation };
glTexEnvi(GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_COMBINE);
glTexEnvi(GL_TEXTURE_ENV, GL_COMBINE_RGB, GL_DOT3_RGB);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE0_RGB, GL_PREVIOUS);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND0_RGB, GL_SRC_COLOR);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE1_RGB, GL_CONSTANT);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND1_RGB, GL_SRC_COLOR);
glTexEnvfv(GL_TEXTURE_ENV, GL_TEXTURE_ENV_COLOR, saturation_constant);
tex->bind();
// Finally we need to interpolate between the original image and the
// greyscale image to get wanted level of saturation
glActiveTexture(GL_TEXTURE2);
glTexEnvi(GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_COMBINE);
glTexEnvi(GL_TEXTURE_ENV, GL_COMBINE_RGB, GL_INTERPOLATE);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE0_RGB, GL_TEXTURE0);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND0_RGB, GL_SRC_COLOR);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE1_RGB, GL_PREVIOUS);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND1_RGB, GL_SRC_COLOR);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE2_RGB, GL_CONSTANT);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND2_RGB, GL_SRC_ALPHA);
glTexEnvfv(GL_TEXTURE_ENV, GL_TEXTURE_ENV_COLOR, saturation_constant);
// Also replace alpha by primary color's alpha here
glTexEnvi(GL_TEXTURE_ENV, GL_COMBINE_ALPHA, GL_REPLACE);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE0_ALPHA, GL_PRIMARY_COLOR);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND0_ALPHA, GL_SRC_ALPHA);
// And make primary color contain the wanted opacity
glColor4f(opacity, opacity, opacity, opacity);
tex->bind();
if (alpha || brightness != 1.0f) {
glActiveTexture(GL_TEXTURE3);
glTexEnvi(GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_COMBINE);
glTexEnvi(GL_TEXTURE_ENV, GL_COMBINE_RGB, GL_MODULATE);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE0_RGB, GL_PREVIOUS);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND0_RGB, GL_SRC_COLOR);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE1_RGB, GL_PRIMARY_COLOR);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND1_RGB, GL_SRC_COLOR);
// The color has to be multiplied by both opacity and brightness
float opacityByBrightness = opacity * brightness;
glColor4f(opacityByBrightness, opacityByBrightness, opacityByBrightness, opacity);
if (alpha) {
// Multiply original texture's alpha by our opacity
glTexEnvi(GL_TEXTURE_ENV, GL_COMBINE_ALPHA, GL_MODULATE);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE0_ALPHA, GL_TEXTURE0);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND0_ALPHA, GL_SRC_ALPHA);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE1_ALPHA, GL_PRIMARY_COLOR);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND1_ALPHA, GL_SRC_ALPHA);
} else {
// Alpha will be taken from previous stage
glTexEnvi(GL_TEXTURE_ENV, GL_COMBINE_ALPHA, GL_REPLACE);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE0_ALPHA, GL_PREVIOUS);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND0_ALPHA, GL_SRC_ALPHA);
}
tex->bind();
}
glActiveTexture(GL_TEXTURE0);
} else if (opacity != 1.0 || brightness != 1.0) {
// the window is additionally configured to have its opacity adjusted,
// do it
float opacityByBrightness = opacity * brightness;
if (alpha) {
glTexEnvi(GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_MODULATE);
glColor4f(opacityByBrightness, opacityByBrightness, opacityByBrightness,
opacity);
} else {
// Multiply color by brightness and replace alpha by opacity
float constant[] = { opacityByBrightness, opacityByBrightness, opacityByBrightness, opacity };
glTexEnvi(GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_COMBINE);
glTexEnvi(GL_TEXTURE_ENV, GL_COMBINE_RGB, GL_MODULATE);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE0_RGB, GL_TEXTURE);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND0_RGB, GL_SRC_COLOR);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE1_RGB, GL_CONSTANT);
glTexEnvi(GL_TEXTURE_ENV, GL_OPERAND1_RGB, GL_SRC_COLOR);
glTexEnvi(GL_TEXTURE_ENV, GL_COMBINE_ALPHA, GL_REPLACE);
glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE0_ALPHA, GL_CONSTANT);
glTexEnvfv(GL_TEXTURE_ENV, GL_TEXTURE_ENV_COLOR, constant);
}
}
#endif
}
void LanczosFilter::restoreRenderStates(GLTexture* tex, double opacity, double brightness, double saturation)
{
#ifdef KWIN_HAVE_OPENGLES
Q_UNUSED(tex)
Q_UNUSED(opacity)
Q_UNUSED(brightness)
Q_UNUSED(saturation)
#else
if (opacity != 1.0 || saturation != 1.0 || brightness != 1.0f) {
if (saturation != 1.0 && tex->saturationSupported()) {
glActiveTexture(GL_TEXTURE3);
glDisable(tex->target());
glActiveTexture(GL_TEXTURE2);
glDisable(tex->target());
glActiveTexture(GL_TEXTURE1);
glDisable(tex->target());
glActiveTexture(GL_TEXTURE0);
}
}
glTexEnvi(GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_REPLACE);
glColor4f(0, 0, 0, 0);
glPopAttrib(); // ENABLE_BIT
#endif
}
/************************************************
* LanczosShader
************************************************/
LanczosShader::LanczosShader(QObject* parent)
: QObject(parent)
, m_shader(0)
, m_uTexUnit(0)
, m_uOffsets(0)
, m_uKernel(0)
, m_arbProgram(0)
{
}
LanczosShader::~LanczosShader()
{
delete m_shader;
#ifndef KWIN_HAVE_OPENGLES
if (m_arbProgram) {
glDeleteProgramsARB(1, &m_arbProgram);
m_arbProgram = 0;
}
#endif
}
void LanczosShader::bind()
{
if (m_shader)
ShaderManager::instance()->pushShader(m_shader);
#ifndef KWIN_HAVE_OPENGLES
else {
glEnable(GL_FRAGMENT_PROGRAM_ARB);
glBindProgramARB(GL_FRAGMENT_PROGRAM_ARB, m_arbProgram);
}
#endif
}
void LanczosShader::unbind()
{
if (m_shader)
ShaderManager::instance()->popShader();
#ifndef KWIN_HAVE_OPENGLES
else {
int boundObject;
glGetProgramivARB(GL_FRAGMENT_PROGRAM_ARB, GL_PROGRAM_BINDING_ARB, &boundObject);
if (boundObject == (int)m_arbProgram) {
glBindProgramARB(GL_FRAGMENT_PROGRAM_ARB, 0);
glDisable(GL_FRAGMENT_PROGRAM_ARB);
}
}
#endif
}
void LanczosShader::setUniforms()
{
if (m_shader) {
glUniform1i(m_uTexUnit, 0);
glUniform2fv(m_uOffsets, 16, (const GLfloat*)m_offsets);
glUniform4fv(m_uKernel, 16, (const GLfloat*)m_kernel);
}
#ifndef KWIN_HAVE_OPENGLES
else {
for (int i = 0; i < 16; ++i) {
glProgramLocalParameter4fARB(GL_FRAGMENT_PROGRAM_ARB, i, m_offsets[i].x(), m_offsets[i].y(), 0, 0);
}
for (int i = 0; i < 16; ++i) {
glProgramLocalParameter4fARB(GL_FRAGMENT_PROGRAM_ARB, i + 16, m_kernel[i].x(), m_kernel[i].y(), m_kernel[i].z(), m_kernel[i].w());
}
}
#endif
}
bool LanczosShader::init()
{
GLPlatform *gl = GLPlatform::instance();
if (gl->supports(GLSL) &&
ShaderManager::instance()->isValid() &&
GLRenderTarget::supported() &&
!(gl->isRadeon() && gl->chipClass() < R600)) {
m_shader = ShaderManager::instance()->loadFragmentShader(ShaderManager::SimpleShader, ":/resources/lanczos-fragment.glsl");
if (m_shader->isValid()) {
ShaderManager::instance()->pushShader(m_shader);
m_uTexUnit = m_shader->uniformLocation("texUnit");
m_uKernel = m_shader->uniformLocation("kernel");
m_uOffsets = m_shader->uniformLocation("offsets");
ShaderManager::instance()->popShader();
return true;
} else {
kDebug(1212) << "Shader is not valid";
m_shader = 0;
// try ARB shader
}
}
#ifdef KWIN_HAVE_OPENGLES
// no ARB shader in GLES
return false;
#else
// try to create an ARB Shader
if (!hasGLExtension("GL_ARB_fragment_program"))
return false;
// We allow Lanczos for SandyBridge or later, but only GLSL shaders are supported, see BUG 301729
if (gl->isIntel())
return false;
QByteArray text;
QTextStream stream(&text);
// Note: This program uses 31 temporaries, 61 ALU instructions, 31 texture
// fetches, 3 texture indirections and 93 instructions.
// The R300 limitations are 32, 64, 32, 4 and 96 respectively.
stream << "!!ARBfp1.0\n";
stream << "TEMP sum;\n";
// Declare 30 temporaries for holding texcoords and TEX results
for (int i = 0; i < 30; i++)
stream << "TEMP temp" << i << ";\n";
// Compute the texture coordinates
for (int i = 0, j = 0; i < 30 / 2; i++) {
stream << "ADD temp" << j++ << ", fragment.texcoord, program.local[" << i + 1 << "];\n";
stream << "SUB temp" << j++ << ", fragment.texcoord, program.local[" << i + 1 << "];\n";
}
// Sample the texture coordinates
stream << "TEX sum, fragment.texcoord, texture[0], 2D;\n";
for (int i = 0; i < 30; i++)
stream << "TEX temp" << i << ", temp" << i << ", texture[0], 2D;\n";
// Process the results
stream << "MUL sum, sum, program.local[16];\n"; // sum = sum * kernel[0]
for (int i = 0, j = 0; i < 30 / 2; i++) {
stream << "MAD sum, temp" << j++ << ", program.local[" << (17 + i) << "], sum;\n";
stream << "MAD sum, temp" << j++ << ", program.local[" << (17 + i) << "], sum;\n";
}
stream << "MOV result.color, sum;\n";
stream << "END\n";
stream.flush();
glEnable(GL_FRAGMENT_PROGRAM_ARB);
glGenProgramsARB(1, &m_arbProgram);
glBindProgramARB(GL_FRAGMENT_PROGRAM_ARB, m_arbProgram);
glProgramStringARB(GL_FRAGMENT_PROGRAM_ARB, GL_PROGRAM_FORMAT_ASCII_ARB, text.length(), text.constData());
if (glGetError()) {
const char *error = (const char*)glGetString(GL_PROGRAM_ERROR_STRING_ARB);
kError() << "Failed to compile fragment program:" << error;
glBindProgramARB(GL_FRAGMENT_PROGRAM_ARB, 0);
glDeleteProgramsARB(1, &m_arbProgram);
glDisable(GL_FRAGMENT_PROGRAM_ARB);
m_arbProgram = 0;
return false;
}
glBindProgramARB(GL_FRAGMENT_PROGRAM_ARB, 0);
glDisable(GL_FRAGMENT_PROGRAM_ARB);
kDebug(1212) << "ARB Shader compiled, id: " << m_arbProgram;
return true;
#endif
}
} // namespace