add SSE path for Matrix4f inverse, taken from Intel except that we do a kosher

division instead of RCPPS-followed-by-Newton-Raphson. The rationale for that is
that elsewhere in Eigen we dont allow ourselves this approximation (which throws
2 bits of mantissa), so there's no reason we should allow it here.
This commit is contained in:
Benoit Jacob
2009-12-14 22:47:14 -05:00
parent d5f3b2dc94
commit d02eccf584
6 changed files with 175 additions and 11 deletions

View File

@@ -25,6 +25,7 @@ namespace Eigen {
#include "src/LU/PartialPivLU.h"
#include "src/LU/Determinant.h"
#include "src/LU/Inverse.h"
#include "src/LU/arch/Inverse_SSE.h"
} // namespace Eigen