mirror of
https://github.com/yuzu-emu/FasTC
synced 2024-11-22 22:03:42 +00:00
a80944901e
In order to better facilitate the change from block stream order to non-block stream order, a lot of changes were introduced to the way that we feed texture data to the compressors. This data is embodied in the CompressionJob struct. We have made it so that the compression job points to both the in and out pointers for our compressed and uncompressed data. Furthermore, we have made sure that the struct also contains the format that its compressing for, so that if any threading programs would like to chop up a compression job into smaller chunks based on the format, it doesn't need to know the format explicitly, it just needs to know certain properties about the format. Moreover, the user can now define the start and end pixels from which we would like to compress to. We can compress subsets of data by changing the in and out pointers and the width and height values. The compressors will read data linearly until they reach the out pixels based on the width of the given pixel.
238 lines
8 KiB
C++
Executable file
238 lines
8 KiB
C++
Executable file
/* FasTC
|
|
* Copyright (c) 2012 University of North Carolina at Chapel Hill. All rights reserved.
|
|
*
|
|
* Permission to use, copy, modify, and distribute this software and its documentation for educational,
|
|
* research, and non-profit purposes, without fee, and without a written agreement is hereby granted,
|
|
* provided that the above copyright notice, this paragraph, and the following four paragraphs appear
|
|
* in all copies.
|
|
*
|
|
* Permission to incorporate this software into commercial products may be obtained by contacting the
|
|
* authors or the Office of Technology Development at the University of North Carolina at Chapel Hill <otd@unc.edu>.
|
|
*
|
|
* This software program and documentation are copyrighted by the University of North Carolina at Chapel Hill.
|
|
* The software program and documentation are supplied "as is," without any accompanying services from the
|
|
* University of North Carolina at Chapel Hill or the authors. The University of North Carolina at Chapel Hill
|
|
* and the authors do not warrant that the operation of the program will be uninterrupted or error-free. The
|
|
* end-user understands that the program was developed for research purposes and is advised not to rely
|
|
* exclusively on the program for any reason.
|
|
*
|
|
* IN NO EVENT SHALL THE UNIVERSITY OF NORTH CAROLINA AT CHAPEL HILL OR THE AUTHORS BE LIABLE TO ANY PARTY FOR
|
|
* DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES, INCLUDING LOST PROFITS, ARISING OUT OF THE
|
|
* USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN IF THE UNIVERSITY OF NORTH CAROLINA AT CHAPEL HILL OR THE
|
|
* AUTHORS HAVE BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
|
*
|
|
* THE UNIVERSITY OF NORTH CAROLINA AT CHAPEL HILL AND THE AUTHORS SPECIFICALLY DISCLAIM ANY WARRANTIES, INCLUDING,
|
|
* BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE AND ANY
|
|
* STATUTORY WARRANTY OF NON-INFRINGEMENT. THE SOFTWARE PROVIDED HEREUNDER IS ON AN "AS IS" BASIS, AND THE UNIVERSITY
|
|
* OF NORTH CAROLINA AT CHAPEL HILL AND THE AUTHORS HAVE NO OBLIGATIONS TO PROVIDE MAINTENANCE, SUPPORT, UPDATES,
|
|
* ENHANCEMENTS, OR MODIFICATIONS.
|
|
*
|
|
* Please send all BUG REPORTS to <pavel@cs.unc.edu>.
|
|
*
|
|
* The authors may be contacted via:
|
|
*
|
|
* Pavel Krajcevski
|
|
* Dept of Computer Science
|
|
* 201 S Columbia St
|
|
* Frederick P. Brooks, Jr. Computer Science Bldg
|
|
* Chapel Hill, NC 27599-3175
|
|
* USA
|
|
*
|
|
* <http://gamma.cs.unc.edu/FasTC/>
|
|
*/
|
|
|
|
#ifndef __COMPRESSION_JOBS_H__
|
|
#define __COMPRESSION_JOBS_H__
|
|
|
|
#include "TexCompTypes.h"
|
|
#include "CompressionFormat.h"
|
|
|
|
#ifdef _MSC_VER
|
|
# define ALIGN(x) __declspec( align(x) )
|
|
#else
|
|
# define ALIGN(x) __attribute__((aligned(x)))
|
|
#endif
|
|
#define ALIGN_SSE ALIGN(16)
|
|
|
|
namespace FasTC {
|
|
|
|
// This structure defines a compression job. Here, width and height are the dimensions
|
|
// of the image in pixels. inBuf contains the R8G8B8A8 data that is to be compressed, and
|
|
// outBuf will contain the compressed BC7 data.
|
|
//
|
|
// Implicit sizes:
|
|
// inBuf - (width * height * 4) bytes
|
|
// outBuf - (width * height) bytes
|
|
class CompressionJob {
|
|
private:
|
|
ECompressionFormat m_Format;
|
|
const uint8 *m_InBuf;
|
|
uint8 *m_OutBuf;
|
|
uint32 m_Width;
|
|
uint32 m_Height;
|
|
uint32 m_XStart, m_XEnd;
|
|
uint32 m_YStart, m_YEnd;
|
|
|
|
public:
|
|
ECompressionFormat Format() const { return m_Format; }
|
|
const uint8 *InBuf() const { return m_InBuf; }
|
|
uint8 *OutBuf() const { return m_OutBuf; }
|
|
uint32 Width() const { return m_Width; }
|
|
uint32 Height() const { return m_Height; }
|
|
uint32 XStart() const { return m_XStart; }
|
|
uint32 XEnd() const { return m_XEnd; }
|
|
uint32 YStart() const { return m_YStart; }
|
|
uint32 YEnd() const { return m_YEnd; }
|
|
|
|
CompressionJob(
|
|
ECompressionFormat _fmt,
|
|
const uint8 *_inBuf,
|
|
unsigned char *_outBuf,
|
|
const uint32 _width,
|
|
const uint32 _height)
|
|
: m_Format(_fmt)
|
|
, m_InBuf(_inBuf)
|
|
, m_OutBuf(_outBuf)
|
|
, m_Width(_width)
|
|
, m_Height(_height)
|
|
, m_XStart(0), m_XEnd(_width)
|
|
, m_YStart(0), m_YEnd(_height)
|
|
{ }
|
|
|
|
CompressionJob(
|
|
ECompressionFormat _fmt,
|
|
const uint8 *_inBuf,
|
|
unsigned char *_outBuf,
|
|
const uint32 _width,
|
|
const uint32 _height,
|
|
const uint32 _xOffset,
|
|
const uint32 _yOffset)
|
|
: m_Format(_fmt)
|
|
, m_InBuf(_inBuf)
|
|
, m_OutBuf(_outBuf)
|
|
, m_Width(_width)
|
|
, m_Height(_height)
|
|
, m_XStart(_xOffset), m_XEnd(_width)
|
|
, m_YStart(_yOffset), m_YEnd(_height)
|
|
{ }
|
|
|
|
CompressionJob(
|
|
ECompressionFormat _fmt,
|
|
const uint8 *_inBuf,
|
|
unsigned char *_outBuf,
|
|
const uint32 _width,
|
|
const uint32 _height,
|
|
const uint32 _xOffset,
|
|
const uint32 _yOffset,
|
|
const uint32 _xEndpoint,
|
|
const uint32 _yEndpoint)
|
|
: m_Format(_fmt)
|
|
, m_InBuf(_inBuf)
|
|
, m_OutBuf(_outBuf)
|
|
, m_Width(_width)
|
|
, m_Height(_height)
|
|
, m_XStart(_xOffset), m_XEnd(_xEndpoint)
|
|
, m_YStart(_yOffset), m_YEnd(_yEndpoint)
|
|
{ }
|
|
|
|
// Returns the x and y coordinates of the pixels that corresponds to the block
|
|
// index for the given format.
|
|
void BlockIdxToCoords(uint32 blockIdx, uint32 (&out)[2]) const {
|
|
uint32 blockDim[2];
|
|
GetBlockDimensions(Format(), blockDim);
|
|
|
|
const uint32 kNumBlocksX = Width() / blockDim[0];
|
|
|
|
const uint32 blockX = blockIdx % kNumBlocksX;
|
|
const uint32 blockY = blockIdx / kNumBlocksX;
|
|
|
|
out[0] = blockX * blockDim[0];
|
|
out[1] = blockY * blockDim[1];
|
|
}
|
|
|
|
// Returns the x and y coordinates of the pixels that corresponds to the block
|
|
// index for the given format.
|
|
uint32 CoordsToBlockIdx(uint32 x, uint32 y) const {
|
|
uint32 blockDim[2];
|
|
GetBlockDimensions(Format(), blockDim);
|
|
|
|
const uint32 kNumBlocksX = Width() / blockDim[0];
|
|
|
|
const uint32 blockX = x / blockDim[0];
|
|
const uint32 blockY = y / blockDim[1];
|
|
|
|
return blockY * kNumBlocksX + blockX;
|
|
}
|
|
};
|
|
|
|
// This struct mirrors that for a compression job, but is used to decompress a BC7 stream. Here, inBuf
|
|
// is a buffer of BC7 data, and outBuf is the destination where we will copy the decompressed R8G8B8A8 data
|
|
class DecompressionJob {
|
|
private:
|
|
const ECompressionFormat m_Format;
|
|
const uint8 *m_InBuf;
|
|
uint8 *m_OutBuf;
|
|
const uint32 m_Width;
|
|
const uint32 m_Height;
|
|
|
|
public:
|
|
const uint8 *InBuf() const { return m_InBuf; }
|
|
uint8 *OutBuf() const { return m_OutBuf; }
|
|
uint32 Width() const { return m_Width; }
|
|
uint32 Height() const { return m_Height; }
|
|
uint32 Format() const { return m_Format; }
|
|
|
|
DecompressionJob(
|
|
ECompressionFormat _fmt,
|
|
const uint8 *_inBuf, uint8 *_outBuf,
|
|
uint32 _width, uint32 _height)
|
|
: m_Format(_fmt)
|
|
, m_InBuf(_inBuf)
|
|
, m_OutBuf(_outBuf)
|
|
, m_Width(_width)
|
|
, m_Height(_height)
|
|
{ }
|
|
};
|
|
|
|
// A structure for maintaining a list of textures to compress.
|
|
class CompressionJobList {
|
|
public:
|
|
|
|
// Initialize the list by specifying the total number of jobs that it will contain.
|
|
// This constructor allocates the necessary memory to hold the array.
|
|
CompressionJobList(const uint32 nJobs);
|
|
~CompressionJobList();
|
|
|
|
// Overrides to deal with memory management.
|
|
CompressionJobList(const CompressionJobList &);
|
|
CompressionJobList &operator =(const CompressionJobList &);
|
|
|
|
// Add a job to the list. This function returns false on failure.
|
|
bool AddJob(const CompressionJob &);
|
|
|
|
// Get the maximum number of jobs that this list can hold.
|
|
uint32 GetTotalNumJobs() const { return m_TotalNumJobs; }
|
|
|
|
// Get the current number of jobs in the list.
|
|
uint32 GetNumJobs() const { return m_NumJobs; }
|
|
|
|
// Returns the specified job.
|
|
const CompressionJob *GetJob(uint32 idx) const;
|
|
uint32 *GetFinishedFlag(uint32 idx) const;
|
|
|
|
private:
|
|
CompressionJob *m_Jobs;
|
|
uint32 m_NumJobs;
|
|
const uint32 m_TotalNumJobs;
|
|
|
|
struct FinishedFlag{
|
|
ALIGN(32) uint32 m_flag;
|
|
} *m_FinishedFlags;
|
|
|
|
public:
|
|
ALIGN(32) uint32 m_CurrentJobIndex;
|
|
ALIGN(32) uint32 m_CurrentBlockIndex;
|
|
};
|
|
|
|
} // namespace FasTC
|
|
#endif // __COMPRESSION_JOBS_H__
|