Files
SingularityViewer/indra/newview/lldrawpoolalpha.cpp
Shyotl 736696ac36 Track glEnable states via static refs instead of map lookups.
Sync light state, bound shader, and various gl context states similarly to render matrices.
Texture handles now refcounted, as multiple viewer textures could ref the same handle (cubemaps do this)
Clean up gl extension loading a bit. Not necessary, but only look for ARB variants if not included in current core version. Removed unused extensions.
Use core shader api if supported, else use ARB. (FN signatures are identical. Just doing some pointer substitution to ARB if not core.)
Attempt at improving VBO update batching. Subdata updates better batched to gether per-frame.
There's probably other stuff I forgot that is in this changeset, too.

Todo: Fix lightstate assertion when toggling fullscreen with shaders off.
2018-11-19 00:37:48 -06:00

591 lines
18 KiB
C++

/**
* @file lldrawpoolalpha.cpp
* @brief LLDrawPoolAlpha class implementation
*
* $LicenseInfo:firstyear=2002&license=viewergpl$
*
* Copyright (c) 2002-2009, Linden Research, Inc.
*
* Second Life Viewer Source Code
* The source code in this file ("Source Code") is provided by Linden Lab
* to you under the terms of the GNU General Public License, version 2.0
* ("GPL"), unless you have obtained a separate licensing agreement
* ("Other License"), formally executed by you and Linden Lab. Terms of
* the GPL can be found in doc/GPL-license.txt in this distribution, or
* online at http://secondlifegrid.net/programs/open_source/licensing/gplv2
*
* There are special exceptions to the terms and conditions of the GPL as
* it is applied to this Source Code. View the full text of the exception
* in the file doc/FLOSS-exception.txt in this software distribution, or
* online at
* http://secondlifegrid.net/programs/open_source/licensing/flossexception
*
* By copying, modifying or distributing this software, you acknowledge
* that you have read and understood your obligations described above,
* and agree to abide by those obligations.
*
* ALL LINDEN LAB SOURCE CODE IS PROVIDED "AS IS." LINDEN LAB MAKES NO
* WARRANTIES, EXPRESS, IMPLIED OR OTHERWISE, REGARDING ITS ACCURACY,
* COMPLETENESS OR PERFORMANCE.
* $/LicenseInfo$
*/
#include "llviewerprecompiledheaders.h"
#include "lldrawpoolalpha.h"
#include "llglheaders.h"
#include "llviewercontrol.h"
#include "llcriticaldamp.h"
#include "llfasttimer.h"
#include "llrender.h"
#include "llcubemap.h"
#include "llsky.h"
#include "lldrawable.h"
#include "llface.h"
#include "llviewercamera.h"
#include "llviewertexturelist.h" // For debugging
#include "llviewerobjectlist.h" // For debugging
#include "llviewerwindow.h"
#include "pipeline.h"
#include "llviewershadermgr.h"
#include "llviewerregion.h"
#include "lldrawpoolwater.h"
#include "llspatialpartition.h"
BOOL LLDrawPoolAlpha::sShowDebugAlpha = FALSE;
static BOOL deferred_render = FALSE;
LLDrawPoolAlpha::LLDrawPoolAlpha(U32 type) :
LLRenderPass(type), current_shader(NULL), target_shader(NULL),
simple_shader(NULL), fullbright_shader(NULL), emissive_shader(NULL),
mColorSFactor(LLRender::BF_UNDEF), mColorDFactor(LLRender::BF_UNDEF),
mAlphaSFactor(LLRender::BF_UNDEF), mAlphaDFactor(LLRender::BF_UNDEF)
{
}
LLDrawPoolAlpha::~LLDrawPoolAlpha()
{
}
void LLDrawPoolAlpha::prerender()
{
mVertexShaderLevel = LLViewerShaderMgr::instance()->getVertexShaderLevel(LLViewerShaderMgr::SHADER_OBJECT);
}
S32 LLDrawPoolAlpha::getNumPostDeferredPasses()
{
static const LLCachedControl<bool> render_depth_of_field("RenderDepthOfField");
if (LLPipeline::sImpostorRender)
{ //skip depth buffer filling pass when rendering impostors
return 1;
}
else if (render_depth_of_field)
{
return 2;
}
else
{
return 1;
}
}
void LLDrawPoolAlpha::beginPostDeferredPass(S32 pass)
{
LL_RECORD_BLOCK_TIME(FTM_RENDER_ALPHA);
if (pass == 0)
{
if (LLPipeline::sImpostorRender)
{
simple_shader = &gDeferredAlphaImpostorProgram;
fullbright_shader = &gDeferredFullbrightProgram;
}
else if (LLPipeline::sUnderWaterRender)
{
simple_shader = &gDeferredAlphaWaterProgram;
fullbright_shader = &gDeferredFullbrightWaterProgram;
}
else
{
simple_shader = &gDeferredAlphaProgram;
fullbright_shader = &gDeferredFullbrightProgram;
}
//F32 gamma = gSavedSettings.getF32("RenderDeferredDisplayGamma");
fullbright_shader->bind();
fullbright_shader->uniform1f(LLShaderMgr::TEXTURE_GAMMA, 2.2f);
//fullbright_shader->uniform1f(LLShaderMgr::DISPLAY_GAMMA, (gamma > 0.1f) ? 1.0f / gamma : (1.0f/2.2f));
fullbright_shader->unbind();
//prime simple shader (loads shadow relevant uniforms)
gPipeline.bindDeferredShader(*simple_shader);
gPipeline.unbindDeferredShader(*simple_shader);
//simple_shader->uniform1f(LLShaderMgr::DISPLAY_GAMMA, (gamma > 0.1f) ? 1.0f / gamma : (1.0f/2.2f));
}
else if (!LLPipeline::sImpostorRender)
{
//update depth buffer sampler
gPipeline.mScreen.flush();
gPipeline.mDeferredDepth.copyContents(gPipeline.mDeferredScreen, 0, 0, gPipeline.mDeferredScreen.getWidth(), gPipeline.mDeferredScreen.getHeight(),
0, 0, gPipeline.mDeferredDepth.getWidth(), gPipeline.mDeferredDepth.getHeight(), GL_DEPTH_BUFFER_BIT, GL_NEAREST);
gPipeline.mDeferredDepth.bindTarget();
simple_shader = fullbright_shader = &gObjectFullbrightProgram[1<<SHD_ALPHA_MASK_BIT];
fullbright_shader->bind();
fullbright_shader->setMinimumAlpha(0.33f);
}
llassert_always(LLPipeline::sRenderDeferred);
emissive_shader = &gDeferredEmissiveProgram;
deferred_render = TRUE;
// Start out with no shaders.
current_shader = target_shader = NULL;
LLGLSLShader::bindNoShader();
}
void LLDrawPoolAlpha::endPostDeferredPass(S32 pass)
{
if (current_shader)
{
gPipeline.unbindDeferredShader(*current_shader);
}
if (pass == 1 && !LLPipeline::sImpostorRender)
{
gPipeline.mDeferredDepth.flush();
gPipeline.mScreen.bindTarget();
gObjectFullbrightProgram[1<<SHD_ALPHA_MASK_BIT].unbind();
}
deferred_render = FALSE;
endRenderPass(pass);
}
void LLDrawPoolAlpha::renderPostDeferred(S32 pass)
{
render(pass);
}
void LLDrawPoolAlpha::beginRenderPass(S32 pass)
{
LL_RECORD_BLOCK_TIME(FTM_RENDER_ALPHA);
simple_shader = &gObjectSimpleProgram[1<<SHD_ALPHA_MASK_BIT | LLPipeline::sUnderWaterRender<<SHD_WATER_BIT];
fullbright_shader = &gObjectFullbrightProgram[1<<SHD_ALPHA_MASK_BIT | LLPipeline::sUnderWaterRender<<SHD_WATER_BIT];
emissive_shader = &gObjectEmissiveProgram[1<<SHD_ALPHA_MASK_BIT | LLPipeline::sUnderWaterRender<<SHD_WATER_BIT];
// Start out with no shaders.
current_shader = target_shader = NULL;
if (mVertexShaderLevel > 0)
{
LLGLSLShader::bindNoShader();
}
}
void LLDrawPoolAlpha::endRenderPass( S32 pass )
{
LL_RECORD_BLOCK_TIME(FTM_RENDER_ALPHA);
LLRenderPass::endRenderPass(pass);
if(mVertexShaderLevel > 0) //Singu Note: Unbind if shaders are enabled at all, not just windlight atmospherics..
{
LLGLSLShader::bindNoShader();
}
}
void LLDrawPoolAlpha::render(S32 pass)
{
LL_RECORD_BLOCK_TIME(FTM_RENDER_ALPHA);
LLGLSPipelineAlpha gls_pipeline_alpha;
LLGLState<GL_LIGHTING> light_state;
gPipeline.enableLightsDynamic(light_state);
if (deferred_render && pass == 1)
{ //depth only
gGL.setColorMask(false, false);
}
else
{
gGL.setColorMask(true, true);
}
bool write_depth = (deferred_render && pass == 1)
// we want depth written so that rendered alpha will
// contribute to the alpha mask used for impostors
|| LLPipeline::sImpostorRenderAlphaDepthPass;
LLGLDepthTest depth(GL_TRUE, write_depth ? GL_TRUE : GL_FALSE);
if (deferred_render && pass == 1)
{
gGL.blendFunc(LLRender::BF_SOURCE_ALPHA, LLRender::BF_ONE_MINUS_SOURCE_ALPHA);
}
else
{
mColorSFactor = LLRender::BF_SOURCE_ALPHA; // } regular alpha blend
mColorDFactor = LLRender::BF_ONE_MINUS_SOURCE_ALPHA; // }
mAlphaSFactor = LLRender::BF_ZERO; // } glow suppression
mAlphaDFactor = LLRender::BF_ONE_MINUS_SOURCE_ALPHA; // }
gGL.blendFunc(mColorSFactor, mColorDFactor, mAlphaSFactor, mAlphaDFactor);
if (mVertexShaderLevel > 0)
{
float min_alpha = LLPipeline::sImpostorRender ? 0.5f : 0.004f;
fullbright_shader->bind();
fullbright_shader->setMinimumAlpha(min_alpha);
simple_shader->bind();
simple_shader->setMinimumAlpha(min_alpha);
emissive_shader->bind();
emissive_shader->setMinimumAlpha(min_alpha);
}
else
{
if (LLPipeline::sImpostorRender)
{
gGL.setAlphaRejectSettings(LLRender::CF_GREATER, 0.5f); //OK
}
else
{
gGL.setAlphaRejectSettings(LLRender::CF_DEFAULT); //OK
}
}
}
renderAlpha(getVertexDataMask(), pass); //getVertexDataMask base mask if fixed-function.
gGL.setColorMask(true, false);
if (deferred_render && pass == 1)
{
gGL.setSceneBlendType(LLRender::BT_ALPHA);
}
if (sShowDebugAlpha)
{
LLGLState<GL_LIGHTING> light_state;
if (LLGLSLShader::sNoFixedFunction)
{
gHighlightProgram.bind();
}
else
{
gPipeline.enableLightsFullbright(light_state);
}
gGL.diffuseColor4f(0.9f,0.f,0.f,0.4f);
LLViewerFetchedTexture::sSmokeImagep->addTextureStats(1024.f*1024.f);
gGL.getTexUnit(0)->bind(LLViewerFetchedTexture::sSmokeImagep, TRUE) ;
renderAlphaHighlight(LLVertexBuffer::MAP_VERTEX |
LLVertexBuffer::MAP_TEXCOORD0);
pushBatches(LLRenderPass::PASS_ALPHA_MASK, LLVertexBuffer::MAP_VERTEX | LLVertexBuffer::MAP_TEXCOORD0, FALSE);
pushBatches(LLRenderPass::PASS_FULLBRIGHT_ALPHA_MASK, LLVertexBuffer::MAP_VERTEX | LLVertexBuffer::MAP_TEXCOORD0, FALSE);
pushBatches(LLRenderPass::PASS_ALPHA_INVISIBLE, LLVertexBuffer::MAP_VERTEX | LLVertexBuffer::MAP_TEXCOORD0, FALSE);
if (LLGLSLShader::sNoFixedFunction)
{
gHighlightProgram.unbind();
}
else
{
gPipeline.enableLightsDynamic(light_state);
}
}
gGL.setSceneBlendType(LLRender::BT_ALPHA);
}
void LLDrawPoolAlpha::renderAlphaHighlight(U32 mask)
{
for (LLCullResult::sg_iterator i = gPipeline.beginAlphaGroups(); i != gPipeline.endAlphaGroups(); ++i)
{
LLSpatialGroup* group = *i;
if (group->getSpatialPartition()->mRenderByGroup &&
!group->isDead())
{
LLSpatialGroup::drawmap_elem_t& draw_info = group->mDrawMap[LLRenderPass::PASS_ALPHA];
for (LLSpatialGroup::drawmap_elem_t::iterator k = draw_info.begin(); k != draw_info.end(); ++k)
{
LLDrawInfo& params = **k;
if (params.mParticle)
{
continue;
}
LLRenderPass::applyModelMatrix(params);
if (params.mGroup)
{
params.mGroup->rebuildMesh();
}
params.mVertexBuffer->setBuffer(mask);
params.mVertexBuffer->drawRange(params.mDrawMode, params.mStart, params.mEnd, params.mCount, params.mOffset);
gPipeline.addTrianglesDrawn(params.mCount, params.mDrawMode);
}
}
}
}
void LLDrawPoolAlpha::renderAlpha(U32 mask, S32 pass)
{
LLGLState<GL_LIGHTING> light_state;
bool light_enabled = TRUE;
bool use_shaders = LLGLSLShader::sNoFixedFunction;
bool depth_only = (pass == 1 && !LLPipeline::sImpostorRender);
for (LLCullResult::sg_iterator i = gPipeline.beginAlphaGroups(); i != gPipeline.endAlphaGroups(); ++i)
{
LLSpatialGroup* group = *i;
llassert(group);
llassert(group->getSpatialPartition());
if (group->getSpatialPartition()->mRenderByGroup &&
!group->isDead())
{
bool is_particle_or_hud_particle = group->getSpatialPartition()->mPartitionType == LLViewerRegion::PARTITION_PARTICLE
|| group->getSpatialPartition()->mPartitionType == LLViewerRegion::PARTITION_CLOUD
|| group->getSpatialPartition()->mPartitionType == LLViewerRegion::PARTITION_HUD_PARTICLE;
bool draw_glow_for_this_partition = !depth_only && mVertexShaderLevel > 0; // no shaders = no glow.
static LLTrace::BlockTimerStatHandle FTM_RENDER_ALPHA_GROUP_LOOP("Alpha Group");
LL_RECORD_BLOCK_TIME(FTM_RENDER_ALPHA_GROUP_LOOP);
bool disable_cull = is_particle_or_hud_particle;
LLGLDisable<GL_CULL_FACE> cull(disable_cull);
LLSpatialGroup::drawmap_elem_t& draw_info = group->mDrawMap[LLRenderPass::PASS_ALPHA];
for (LLSpatialGroup::drawmap_elem_t::iterator k = draw_info.begin(); k != draw_info.end(); ++k)
{
LLDrawInfo& params = **k;
/*if ((params.mVertexBuffer->getTypeMask() & mask) != mask)
{ //FIXME!
LL_WARNS() << "Missing required components, skipping render batch." << LL_ENDL;
continue;
}*/
// Fix for bug - NORSPEC-271
// If the face is more than 90% transparent, then don't update the Depth buffer for Dof
// We don't want the nearly invisible objects to cause of DoF effects
if(depth_only)
{
LLFace* face = params.mFace;
if(face)
{
const LLTextureEntry* tep = face->getTextureEntry();
if(tep)
{
if(tep->getColor().mV[3] < 0.1f)
continue;
}
}
}
LLRenderPass::applyModelMatrix(params);
// Singu Note: Logic reordered here. Removed a fair bit of unneeded condition checks (mostly related to deferred mode), and made
// shader selection more explicit.
if (params.mGroup)
{
params.mGroup->rebuildMesh();
}
LLMaterial* mat = deferred_render ? params.mMaterial.get() : NULL;
if (!use_shaders)
{
llassert_always(!target_shader);
llassert_always(!current_shader);
llassert_always(!LLGLSLShader::sNoFixedFunction);
llassert_always(!LLGLSLShader::sCurBoundShaderPtr);
bool fullbright = depth_only || params.mFullbright;
if(light_enabled == fullbright)
{
light_enabled = !fullbright;
if (light_enabled) // Set local lights
{
gPipeline.enableLightsDynamic(light_state);
}
else // Set fullbright
{
gPipeline.enableLightsFullbright(light_state);
}
}
}
else
{
if (mat)
{
U32 mask = params.mShaderMask;
llassert(mask < LLMaterial::SHADER_COUNT);
target_shader = LLPipeline::sUnderWaterRender ? &(gDeferredMaterialWaterProgram[mask]) : &(gDeferredMaterialProgram[mask]);
// If we need shaders, and we're not ALREADY using the proper shader, then bind it
// (this way we won't rebind shaders unnecessarily).
if (current_shader != target_shader)
{
if(current_shader)
gPipeline.unbindDeferredShader(*current_shader);
gPipeline.bindDeferredShader(*target_shader);
}
}
else
{
target_shader = params.mFullbright ? fullbright_shader : simple_shader;
// If we need shaders, and we're not ALREADY using the proper shader, then bind it
// (this way we won't rebind shaders unnecessarily).
if(current_shader != target_shader)
{
if(current_shader)
gPipeline.unbindDeferredShader(*current_shader);
gPipeline.bindDeferredShader(*target_shader);
}
}
current_shader = target_shader;
if(mat)
{
current_shader->uniform4f(LLShaderMgr::SPECULAR_COLOR, params.mSpecColor.mV[0], params.mSpecColor.mV[1], params.mSpecColor.mV[2], params.mSpecColor.mV[3]);
current_shader->uniform1f(LLShaderMgr::ENVIRONMENT_INTENSITY, params.mEnvIntensity);
current_shader->uniform1f(LLShaderMgr::EMISSIVE_BRIGHTNESS, params.mFullbright ? 1.f : 0.f);
if (params.mNormalMap)
{
params.mNormalMap->addTextureStats(params.mVSize);
current_shader->bindTexture(LLShaderMgr::BUMP_MAP, params.mNormalMap);
}
if (params.mSpecularMap)
{
params.mSpecularMap->addTextureStats(params.mVSize);
current_shader->bindTexture(LLShaderMgr::SPECULAR_MAP, params.mSpecularMap);
}
}
/*else if(deferred_render && current_shader == simple_shader)
{
current_shader->uniform4f(LLShaderMgr::SPECULAR_COLOR, 1.0f, 1.0f, 1.0f, 1.0f);
current_shader->uniform1f(LLShaderMgr::ENVIRONMENT_INTENSITY, 0.0f);
LLViewerFetchedTexture::sFlatNormalImagep->addTextureStats(params.mVSize);
current_shader->bindTexture(LLShaderMgr::BUMP_MAP, LLViewerFetchedTexture::sFlatNormalImagep);
LLViewerFetchedTexture::sWhiteImagep->addTextureStats(params.mVSize);
current_shader->bindTexture(LLShaderMgr::SPECULAR_MAP, LLViewerFetchedTexture::sWhiteImagep);
}*/
if (params.mTextureList.size() > 1)
{
for (U32 i = 0; i < params.mTextureList.size(); ++i)
{
if (params.mTextureList[i].notNull())
{
gGL.getTexUnit(i)->bind(params.mTextureList[i], TRUE);
}
}
}
}
bool tex_setup = false;
if(!use_shaders || params.mTextureList.size() <= 1)
{ //not batching textures or batch has only 1 texture -- might need a texture matrix
if (params.mTexture.notNull())
{
params.mTexture->addTextureStats(params.mVSize);
if (use_shaders && mat && current_shader)
{
current_shader->bindTexture(LLShaderMgr::DIFFUSE_MAP, params.mTexture);
}
else
{
gGL.getTexUnit(0)->bind(params.mTexture, TRUE);
}
if (params.mTextureMatrix)
{
tex_setup = true;
gGL.getTexUnit(0)->activate();
gGL.matrixMode(LLRender::MM_TEXTURE);
gGL.loadMatrix(*params.mTextureMatrix);
gPipeline.mTextureMatrixOps++;
}
}
else
{
gGL.getTexUnit(0)->unbind(LLTexUnit::TT_TEXTURE);
}
}
static LLTrace::BlockTimerStatHandle FTM_RENDER_ALPHA_PUSH("Alpha Push Verts");
{
LL_RECORD_BLOCK_TIME(FTM_RENDER_ALPHA_PUSH);
gGL.blendFunc((LLRender::eBlendFactor) params.mBlendFuncSrc, (LLRender::eBlendFactor) params.mBlendFuncDst, mAlphaSFactor, mAlphaDFactor);
// Singu Note: If using shaders, pull the attribute mask from it, else used passed base mask.
params.mVertexBuffer->setBuffer(current_shader ? current_shader->mAttributeMask : mask);
params.mVertexBuffer->drawRange(params.mDrawMode, params.mStart, params.mEnd, params.mCount, params.mOffset);
gPipeline.addTrianglesDrawn(params.mCount, params.mDrawMode);
}
// If this alpha mesh has glow, then draw it a second time to add the destination-alpha (=glow). Interleaving these state-changing calls could be expensive, but glow must be drawn Z-sorted with alpha.
if (current_shader &&
draw_glow_for_this_partition &&
(!is_particle_or_hud_particle || params.mHasGlow) && //only do this second pass for batches that actually have glow
params.mVertexBuffer->hasDataType(LLVertexBuffer::TYPE_EMISSIVE))
{
// install glow-accumulating blend mode
gGL.blendFunc(LLRender::BF_ZERO, LLRender::BF_ONE, // don't touch color
LLRender::BF_ONE, LLRender::BF_ONE); // add to alpha (glow)
//emissive_shader->bind();
// glow doesn't use vertex colors from the mesh data
// Singu Note: Pull attribs from shader, since we always have one here.
// Singu Note: To avoid ridiculous shader bind cost, simply re-use prior shader, but let llvertexbuffer replace the color attrib ptr with the emissive one.
params.mVertexBuffer->setBuffer(current_shader->mAttributeMask | LLVertexBuffer::MAP_EMISSIVE);
// do the actual drawing, again
params.mVertexBuffer->drawRange(params.mDrawMode, params.mStart, params.mEnd, params.mCount, params.mOffset);
gPipeline.addTrianglesDrawn(params.mCount, params.mDrawMode);
//current_shader->bind();
// restore our alpha blend mode
gGL.blendFunc(mColorSFactor, mColorDFactor, mAlphaSFactor, mAlphaDFactor);
}
if (tex_setup)
{
gGL.getTexUnit(0)->activate();
gGL.loadIdentity();
gGL.matrixMode(LLRender::MM_MODELVIEW);
}
}
}
}
if (!light_enabled)
{
gPipeline.enableLightsDynamic(light_state);
}
gGL.setSceneBlendType(LLRender::BT_ALPHA);
LLVertexBuffer::unbind();
}