mi32/cpuext.h File Reference
#include <mi32/stddefns.h>#include <stddef.h>
Go to the source code of this file.
Defines | |
| #define | amemcpy memcpy |
| #define | CPUEXT_3DNOW 0x00010000 |
| #define | CPUEXT_3DNOWEX 0x00020000 |
| #define | CPUEXT_AMDMMX 0x00040000 |
| #define | CPUEXT_CMOV 0x00000010 |
| #define | CPUEXT_MMX 0x00000001 |
| #define | CPUEXT_PREFETCHNTA 0x00000004 |
| #define | CPUEXT_SIMD 0x00000002 |
| #define | CPUEXT_SSE 0x00000002 |
| #define | CPUEXT_SSE2 0x00000008 |
| #define | CPUEXT_TSC 0x00000020 |
| #define | GetCPUExtension() (0) |
| #define | HasMMX() (GetCPUExtension() & CPUEXT_MMX) |
| #define | LIBEXPORT MI_DLLIMPORT |
Functions | |
| void | AlphaBlend24 (UINT8 *dest, const UINT8 *src, UINT8 alpha, int n) |
| void | AlphaBlend32 (UINT32 *dest, const UINT32 *src, UINT8 alpha, int n) |
| void | AlphaBlendColor24 (UINT8 *dest, UINT8 r, UINT8 g, UINT8 b, UINT8 alpha, int n) |
| void | AlphaBlendColor32 (UINT32 *dest, UINT32 pix, UINT8 alpha, int n) |
| void | AlphaBlendColor32ABGR (UINT32 *dest, UINT32 pix, int n) |
| void | AlphaBlendColor32RGBA (UINT32 *dest, UINT32 pix, int n) |
| void | AlphaBlendVarAlpha24 (UINT8 *dest, const UINT8 *src, const UINT8 *alpha, int n) |
| void | AlphaBlendVarAlpha32 (UINT32 *dest, const UINT32 *src, const UINT8 *alpha, int n) |
Define Documentation
| #define amemcpy memcpy |
Intel.
| #define CPUEXT_3DNOW 0x00010000 |
AMD 3DNow.
| #define CPUEXT_3DNOWEX 0x00020000 |
AMD 3DNow extensions.
| #define CPUEXT_AMDMMX 0x00040000 |
AMD's MMX extensions.
| #define CPUEXT_CMOV 0x00000010 |
| #define CPUEXT_MMX 0x00000001 |
Basic MMX support.
| #define CPUEXT_PREFETCHNTA 0x00000004 |
Intel cashability instructions.
| #define CPUEXT_SIMD 0x00000002 |
Old name for SSE.
| #define CPUEXT_SSE 0x00000002 |
Old name for SSE.
| #define CPUEXT_SSE2 0x00000008 |
| #define CPUEXT_TSC 0x00000020 |
| #define GetCPUExtension | ( | ) | (0) |
< Not a chance!
| #define HasMMX | ( | ) | (GetCPUExtension() & CPUEXT_MMX) |
| #define LIBEXPORT MI_DLLIMPORT |
Function Documentation
< End of fake functions
Do alpha blending of a source buffer into a dest buffer. This version is for 24-bit source onto 24-bit dest. n is number of pixels. alpha is in the range 0-255 (although the extremes are rather pointless, as 0 would leave the dest untouched and 255 is an expensive way to do a memcpy)
On platforms where it's possible, this function checks for the availability of MMX and uses it if it's there
Do alpha blending of a source buffer into a dest buffer.
This version is for 32-bit source onto 32-bit dest. n is number of pixels. alpha is in the range 0-255 (although the extremes are rather pointless, as 0 would leave the dest untouched and 255 is an expensive way to do a memcpy)
This version assumes a constant alpha. Note that all 4 color componants are blended because it's faster to do that (4 bytes at a time) than doing 3 individual componants and skipping the 4th. It also means this function doesn't have to care what the RGB order is and which byte to skip.
On platforms where it's possible, this function checks for the availability of MMX and uses it if it's there
Do alpha blending of a solid color onto a destination buffer.
The color is broken up into 3 componants because the places where this will be the most useful, we already have them like this. implemented in aalphab.c Dest is assumed to be 24-bit RGB data. If it's really BGR, pass your color componants backwards.
On platforms where it's possible, this function checks for the availability of MMX and uses it if it's there
Do alpha blending of a solid color onto a destination buffer.
implemented in aalphab.c
Note that all 4 color componants are blended because it's faster to do that (4 bytes at a time) than doing 3 individual componants and skipping the 4th. It also means this function doesn't have to care what the RGB order is and which byte to skip.
On platforms where it's possible, this function checks for the availability of MMX and uses it if it's there
Do alpha blending of a solid color onto a destination buffer.
implemented in aalphab.c
This is similar to AlphaBlendColor24() but the dest buffer is assumed to be made up of 32-bit pixels in which the highet 8 bits are "alpha". Note that this is not based on the order of the bytes in the buffer, but on bits in a 32-bit integer. It will work for 0xAABBGGRR or 0xAARRGGBB. Also note that for this one, the "alpha" value is part of the pixel passed in.
Do alpha blending of a solid color onto a destination buffer.
implemented in aalphab.c
This is similar to AlphaBlendColor24() but the dest buffer is assumed to be made up of 32-bit pixels in which the lowest 8 bits are "alpha". Note that this is not based on the order of the bytes in the buffer, but on bits in a 32-bit integer. It will work for 0xRRGGBBAA or 0xBBGGRRAA. Also note that for this one, the "alpha" value is part of the pixel passed in.
Do alpha blending of a source buffer into a dest buffer.
This version is for 24-bit source onto 24-bit dest. n is number of pixels. alpha is in the range 0-255
On platforms where it's possible, this function checks for the availability of MMX and uses it if it's there
- Parameters:
-
alpha Alphas for each pixel.
Do alpha blending of a source buffer into a dest buffer.
This version is for 32-bit source onto 32-bit dest. n is number of pixels. alpha is in the range 0-255
Note that all 4 color componants are blended because it's faster to do that (4 bytes at a time) than doing 3 individual componants and skipping the 4th. It also means this function doesn't have to care what the RGB order is and which byte to skip.
On platforms where it's possible, this function checks for the availability of MMX and uses it if it's there
