Another selb attempt
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| # Rough sketch, not actually tested or anything... | |
| # in any case, way too many instrs to be actually useful. | |
| rotqbyi inMaskY, inMask, 4 | |
| rotqbyi inMaskZ, inMask, 8 | |
| rotqbyi inMaskW, inMask, 12 | |
| cwd maskX, 4(inMask) # 0x00010203 in maskX.x if inMask.x set, 0x10111213 otherwise (rest id + 0x10) | |
| cwd maskY, 8(inMaskY) # 0x00010203 in maskY.y if inMask.y set, 0x14151617 otherwise (rest id + 0x10) | |
| cwd maskZ, 12(inMaskZ) # 0x00010203 in maskZ.z if inMask.z set, 0x18191a1b otherwise (rest id + 0x10) | |
| cwd maskW, 0(inMaskW) # 0x00010203 in maskZ.w if inMask.w set, 0x1c1d1e1f otherwise (rest id + 0x10) | |
| # magicY.x = 0x04050607 | |
| # magicZ.x = 0x08090a0b | |
| # magicW.x = 0x0c0d0e0f | |
| shufb outY, magicY, maskX, maskY | |
| shufb outZ, magicZ, outY, maskZ | |
| shufb outW, magicW, outZ, maskW | |
| # now outW contains shufb mask (but positions of a and b reversed!) | |
| shufb result, val_b, val_a, outW |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment