The trick consists in introducing a new pload1 function to be used in low level product kernels for which bug #203 does not apply.
Indeed, it turned out that using inline assembly prevents gcc of doing a good job at instructtion reordering.
- remove most of the metaprogramming kung fu in MathFunctions.h (only keep functions that differs from the std)
- remove the overloads for array expression that were in the std namespace
To get the links to work, I also had to document the Eigen namespace.
Unfortunately, this means that the word Eigen is linked whenever it appears
in the docs.
* pload* and pset1 are now templated on the packet type
* gemv routines are now embeded into a structure with
a consistent API with respect to gemm
* some configurations of vector * matrix and matrix * matrix works fine,
some need more work...
replaced _mm_prefetch in GeneralBlockPanelKernel.h, with ei_prefetch() inline function.
Implemented NEON and AltiVec versions, copied SSE version over from GeneralBlockPanelKernel.h.
Also in GCC case (or rather !_MSC_VER) it's implemented using __builtin_prefetch().
NEON managed to give a small but welcome boost, 0.88GFLOPS -> 0.91GFLOPS.
Implemented NEON and AltiVec versions, copied SSE version over from GeneralBlockPanelKernel.h.
Also in GCC case (or rather !_MSC_VER) it's implemented using __builtin_prefetch().
NEON managed to give a small but welcome boost, 0.88GFLOPS -> 0.91GFLOPS.
Added setLinSpaced/LinSpaced functionality to DenseBase.
Improved vectorized assignment - overcomes MSVC optimization issues.
CwiseNullaryOp is now requiring functors to offer 1D and 2D operators.
Adapted existing functors to the new CwiseNullaryOp requirements.
Added ei_plset to create packages as [a, a+1, ..., a+size].
Added more nullaray unit tests.
Add an internal pseudo expression allowing to optimize operators like +=, *= using
the copyCoeff stuff.
This allows to easily enforce aligned load for the destination matrix everywhere.
Pommier. They are for float only, and they return exactly the same
result as the standard versions in about 90% of the cases. Otherwise the max error
is below 1e-7. However, for very large values (>1e3) the accuracy of sin and cos
slighlty decrease. They are about 3 or 4 times faster than 4 calls to their respective
standard versions. So, is it ok to enable them by default in their respective functors ?
