0ff1026dbb
Removed value_type and size_type
ago%!(EXTRA string=12 years)
b9266c07b0
Using C++ casts
ago%!(EXTRA string=12 years)
16b9c79acc
Added implicit conversion of types with different precision qualifier
ago%!(EXTRA string=12 years)
25a5c21a24
Initial build with half types removed
ago%!(EXTRA string=12 years)
ef15d06104
Updated doxygen references
ago%!(EXTRA string=12 years)
0a54eb99b1
Fixed compatibility type, using precision qualifier
ago%!(EXTRA string=12 years)
1b3d5cd378
Fixed space characters
ago%!(EXTRA string=12 years)
a319cff426
Removed scalar functions
ago%!(EXTRA string=12 years)
691f04e14f
Fixed post in/decrement operators
ago%!(EXTRA string=12 years)
b2e55904b8
Fixed dual quat built on GCC
ago%!(EXTRA string=12 years)
13f27821b9
Completed GTX_multiple for negative values, issue #79
ago%!(EXTRA string=12 years)
f0d4080fe3
Fix glm::higherMultiple and glm::lowerMultiple and assoicated test cases. Does not fix the specialized versions of these functions
ago%!(EXTRA string=12 years)
e0cfd7d672
Attempt a small optimization in operator*(fquatSIMD, fvec4SIMD). No big improvement.
ago%!(EXTRA string=12 years)
c08ea7656b
Reorganize operator* to makes things a bit clearer.
ago%!(EXTRA string=12 years)
1eb88e4bc1
Cleanup.
ago%!(EXTRA string=12 years)
c1006718b3
Add fastMix() and fastSlerp() implementations.
...
These have stricter pre-conditions than standard mix() and slerp()
- 1) Input quaternions must be unit length.
- 2) The interpolation factor (a) must be in the range [0, 1]
None of these restrictions should be too bad. The reason for these is that it uses fastAcos()
and fastSin(), both of which have a limited allowable range.
In my contrived tests, I observed about a 10x improvement over the standard versions. This is
mostly because of the faster acos/sin operations. The fastSin(__m128) implementation also helps
here because it can do four fastSin() operations simultaneously using SSE (mix() and slerp()
each need three).
ago%!(EXTRA string=12 years)
d07496460a
Add the ability to convert a SIMD quat to a standard mat4.
ago%!(EXTRA string=12 years)
942bf08fe3
Add the ability to convert standard mat4s and mat3s to SIMD quats.
ago%!(EXTRA string=12 years)
13837e1079
Remove the SSE3 implementation in operator*(fquatSIMD, fquatSIMD).
...
The SSE2 version is now running faster than the SSE3 version.
ago%!(EXTRA string=12 years)
197b6c96d8
Improve efficiency of operator*(fquatSIMD, fquatSIMD) in SSE4 mode.
...
Now only requires 3 shuffle, 4 mul and 4 dpps.
ago%!(EXTRA string=12 years)
8c7828e6f7
Fixed merge
ago%!(EXTRA string=12 years)
3c7989486e
Fixed literals in dual quaternion implementation
ago%!(EXTRA string=12 years)
7563a8bc4d
Add initial implementation of SIMD optimized quaternions.
...
A few things here can probably be improved by people a lot smarter then
me, but for the most part things are generally faster.
A few notes:
- A fquatSIMD can be converted to a fmat4x4SIMD using mat4SIMD_cast().
- A tquat<float> can be converted to a fquatSIMD using quatSIMD_cast().
- Some functions are virtually the same as their scalar counterparts
because I've just not been able to get them faster.
- Only the basic functions are implemented. Future plans include fast,
approximate normalize, length and mix/slerp functions.
ago%!(EXTRA string=12 years)
6a7ccdb530
Fix another compilation error.
ago%!(EXTRA string=12 years)
b6d994f638
Compilation fixes due to missing P template parameters.
ago%!(EXTRA string=12 years)
5519b86003
Fix fvec4SIMD and fmat4x4SIMD compilation errors.
ago%!(EXTRA string=12 years)
f1967b5fb3
Removed GTX_ocl_type, issue #64
ago%!(EXTRA string=12 years)
e72f18997c
Fixed GTX_multiple for negative values, issue #51
ago%!(EXTRA string=12 years)
d4043ea49a
Added all precision types for GTC_type_precision
ago%!(EXTRA string=12 years)
296e3d7007
Fixed half precision implementation
ago%!(EXTRA string=12 years)
c14e2d7fbc
Added precision template parameter
ago%!(EXTRA string=12 years)
39cf417691
Added initial implementation for forward declarations: int and float scalar types, #56
ago%!(EXTRA string=13 years)
6a96cd819a
Fixed dual quaternion build on Clang, fixed space characters #49
ago%!(EXTRA string=13 years)
78d3e5ca66
Updated readme
ago%!(EXTRA string=13 years)
8ed5f541ff
Resolved issue #58
ago%!(EXTRA string=13 years)
4d71624067
Added file headers so that the files are no longer empty. #57
ago%!(EXTRA string=13 years)
6c4ae0d241
Fixed space characters
ago%!(EXTRA string=13 years)
f37af51426
Fixed API for euclian, only need a vec2 parameter
ago%!(EXTRA string=13 years)
982bb5ce5d
Added rotation function, issue #22
ago%!(EXTRA string=13 years)
18a8efde62
Resolve issue #39 , exp, squad and intermediate functions
ago%!(EXTRA string=13 years)
1e3cb00fe5
Completed bit interleaving for 3 and 4 values
ago%!(EXTRA string=13 years)
3e3736769c
Fixed intrinsics build
ago%!(EXTRA string=13 years)
2a4434786e
Updated date
ago%!(EXTRA string=13 years)
50c870ea96
Moved all dual quaternion functionality from GTC space to GTX.
ago%!(EXTRA string=13 years)
bd7125c50b
Added bit interleave for 3 and 4 integers
ago%!(EXTRA string=13 years)
6799127ec6
Fixed any and all for scalars
ago%!(EXTRA string=13 years)
59b9a780e9
Added GTX_scalar_relational implementation and tests
ago%!(EXTRA string=13 years)
c57a000791
Added bitfieldInterleave implementation
ago%!(EXTRA string=13 years)
920ca0a242
Experiments to add bitfieldInterleave
ago%!(EXTRA string=13 years)
a4aa39f9bd
Fixed 0x2013 dash character in comments that cause issue in Windows
ago%!(EXTRA string=13 years)
Christophe Riccio
Christophe Riccio
Robert Ancell
Dave Reid
Maksim Vorobiev