Optimize outer products for non rank-1 update operation

Submitted by Gael Guennebaud @ggael

Assigned to Nobody

Link to original bugzilla bug (#483)

Description

Currently the outer product impl. is optimized for:

A += v1 * v2.transpose();

making a simple:

A = v1 * v2.transpose();

significantly slower than what it could be because of the additional A.setZero() and additional adds.

Blocking

#387 (closed)