Mirrors/overte-JulianGro
3
0
Fork 0
mirror of https://github.com/JulianGro/overte.git synced 2025-04-25 17:14:59 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

Adjust gl45 backend to request interest in mips and keep track of min requested

Browse source
This commit is contained in:
Ryan Huffman 2017-04-21 01:11:30 -07:00 committed by Atlante45
parent 970be9d2c5
commit 14f8c91e23
4 changed files with 17 additions and 11 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
libraries/gpu-gl/src/gpu/gl45/GL45Backend.h
View file

@ -117,6 +117,8 @@ public:
void allocateStorage(uint16 mip);
void copyMipsFromTexture();
uint16 _lowestRequestedMip { 0 };
};
#if 0

2
libraries/gpu-gl/src/gpu/gl45/GL45BackendTexture.cpp
View file

@ -119,8 +119,10 @@ GL45Texture::GL45Texture(const std::weak_ptr<GLBackend>& backend, const Texture&
GLuint GL45Texture::allocate(const Texture& texture) {
GLuint result;
glCreateTextures(getGLTextureType(texture), 1, &result);
#ifdef DEBUG
auto source = texture.source();
glObjectLabel(GL_TEXTURE, result, source.length(), source.data());
#endif
return result;
}

23
libraries/gpu-gl/src/gpu/gl45/GL45BackendVariableTexture.cpp
View file

@ -75,6 +75,11 @@ void GL45ResourceTexture::allocateStorage(uint16 allocatedMip) {
for (uint16_t mip = _allocatedMip; mip < mipLevels; ++mip) {
_size += _gpuObject.evalMipSize(mip);
}
if (!_gpuObject.isStoredMipFaceAvailable(allocatedMip, 0)) {
const_cast<gpu::Texture&>(_gpuObject).requestInterestInMip(allocatedMip);
}
Backend::updateTextureGPUMemoryUsage(0, _size);
}
@ -98,9 +103,11 @@ void GL45ResourceTexture::syncSampler() const {
void GL45ResourceTexture::promote() {
PROFILE_RANGE(render_gpu_gl, __FUNCTION__);
//Q_ASSERT(_allocatedMip > 0);
if (!_gpuObject.isStoredMipFaceAvailable(_populatedMip - 1, 0)) {
const_cast<gpu::Texture&>(_gpuObject).requestInterestInMip(_populatedMip - 1);
Q_ASSERT(_allocatedMip > 0);
uint16_t targetAllocatedMip = _allocatedMip - std::min<uint16_t>(_allocatedMip, 2);
if (!_gpuObject.isStoredMipFaceAvailable(targetAllocatedMip, 0)) {
_lowestRequestedMip = targetAllocatedMip;
const_cast<gpu::Texture&>(_gpuObject).requestInterestInMip(targetAllocatedMip);
return;
}
GLuint oldId = _id;
@ -109,7 +116,7 @@ void GL45ResourceTexture::promote() {
const_cast<GLuint&>(_id) = allocate(_gpuObject);
uint16_t oldAllocatedMip = _allocatedMip;
// allocate storage for new level
allocateStorage(_allocatedMip - std::min<uint16_t>(_allocatedMip, 2));
allocateStorage(targetAllocatedMip);
uint16_t mips = _gpuObject.getNumMips();
// copy pre-existing mips
for (uint16_t mip = _populatedMip; mip < mips; ++mip) {
@ -176,23 +183,21 @@ void GL45ResourceTexture::populateTransferQueue() {
const uint8_t maxFace = GLTexture::getFaceCount(_target);
uint16_t sourceMip = _populatedMip;
//qDebug() << "populateTransferQueue info : " << _populatedMip << " " << _maxAllocatedMip << " " << _allocatedMip;
do {
--sourceMip;
auto targetMip = sourceMip - _allocatedMip;
auto mipDimensions = _gpuObject.evalMipDimensions(sourceMip);
bool transferQueued = false;
//qDebug() << "populateTransferQueue " << QString::fromStdString(_gpuObject.source()) << sourceMip << " " << targetMip;
for (uint8_t face = 0; face < maxFace; ++face) {
if (!_gpuObject.isStoredMipFaceAvailable(sourceMip, face)) {
const_cast<gpu::Texture&>(_gpuObject).requestInterestInMip(sourceMip);
_minRequestedMip = sourceMip;
_lowestRequestedMip = sourceMip;
continue;
}
_lowestRequestedMip = sourceMip;
// If the mip is less than the max transfer size, then just do it in one transfer
if (glm::all(glm::lessThanEqual(mipDimensions, MAX_TRANSFER_DIMENSIONS))) {
qDebug() << "mip is less than max transfer size";
// Can the mip be transferred in one go
_pendingTransfers.emplace(new TransferJob(*this, sourceMip, targetMip, face));
transferQueued = true;
@ -207,14 +212,12 @@ void GL45ResourceTexture::populateTransferQueue() {
Q_ASSERT(0 == (mipSize % lines));
uint32_t linesPerTransfer = (uint32_t)(MAX_TRANSFER_SIZE / bytesPerLine);
uint32_t lineOffset = 0;
qDebug() << "queing up single line transfers " << linesPerTransfer << " " << lineOffset;
while (lineOffset < lines) {
uint32_t linesToCopy = std::min<uint32_t>(lines - lineOffset, linesPerTransfer);
_pendingTransfers.emplace(new TransferJob(*this, sourceMip, targetMip, face, linesToCopy, lineOffset));
lineOffset += linesToCopy;
transferQueued = true;
}
_minRequestedMip = std::min(_minRequestedMip, sourceMip);
}
// queue up the sampler and populated mip change for after the transfer has completed

1
libraries/gpu/src/gpu/Texture.cpp
View file

@ -397,7 +397,6 @@ Size Texture::evalTotalSize(uint16 startingMip) const {
Size size = 0;
uint16 minMipLevel = std::max(getMinMip(), startingMip);
uint16 maxMipLevel = getMaxMip();
qDebug() << " min: " << minMipLevel << " " << maxMipLevel;
for (uint16 level = minMipLevel; level <= maxMipLevel; level++) {
size += evalMipSize(level);
}